3bad2cd6e16f6b3bec0eba503322d26629dd16e7
[WebKit-https.git] / Tools / ChangeLog
1 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
2
3         nrwt: merge WebKitDriver into Driver
4         https://bugs.webkit.org/show_bug.cgi?id=91696
5
6         Reviewed by Adam Barth.
7
8         As part of merging WebKitPort into Port, we also should merge
9         WebKitDriver into Driver, since all of the non-test Drivers now
10         inherit from WebKitDriver.
11
12         There are no functional changes in this patch, and all the changes should
13         be covered by the existing tests.
14
15         * Scripts/webkitpy/layout_tests/port/base.py:
16         (Port._driver_class):
17         * Scripts/webkitpy/layout_tests/port/chromium.py:
18         (ChromiumPort._check_file_exists):
19         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
20         (ChromiumAndroidDriver):
21         (ChromiumAndroidDriver.__init__):
22         (ChromiumAndroidDriver._drt_cmd_line):
23         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
24         (ChromiumAndroidDriverTest.test_read_prompt):
25         * Scripts/webkitpy/layout_tests/port/driver.py:
26         (Driver):
27         (Driver.__init__):
28         (Driver.__del__):
29         (Driver.run_test):
30         (Driver.has_crashed):
31         (Driver.start):
32         (Driver._start):
33         (Driver.stop):
34         (Driver.cmd_line):
35         (Driver._check_for_driver_crash):
36         (Driver._command_from_driver_input):
37         (Driver._read_first_block):
38         (Driver._read_optional_image_block):
39         (Driver._read_header):
40         (Driver._process_stdout_line):
41         (Driver._strip_eof):
42         (Driver._read_block):
43         (ContentBlock):
44         (ContentBlock.__init__):
45         (ContentBlock.decode_content):
46         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
47         (DriverTest.make_port):
48         (DriverTest._assert_wrapper):
49         (DriverTest.test_read_block):
50         (DriverTest):
51         (DriverTest.test_read_binary_block):
52         (DriverTest.test_read_base64_block):
53         (DriverTest.test_no_timeout):
54         (DriverTest.test_check_for_driver_crash):
55         (DriverTest.test_check_for_driver_crash.FakeServerProcess):
56         (DriverTest.test_check_for_driver_crash.FakeServerProcess.__init__):
57         (DriverTest.test_check_for_driver_crash.FakeServerProcess.pid):
58         (DriverTest.test_check_for_driver_crash.FakeServerProcess.name):
59         (DriverTest.test_check_for_driver_crash.FakeServerProcess.has_crashed):
60         (DriverTest.test_check_for_driver_crash.FakeServerProcess.stop):
61         (DriverTest.test_check_for_driver_crash.assert_crash):
62         (DriverTest.test_creating_a_port_does_not_write_to_the_filesystem):
63         (DriverTest.test_stop_cleans_up_properly):
64         (DriverTest.test_two_starts_cleans_up_properly):
65         (MockServerProcess):
66         (MockServerProcess.__init__):
67         (MockServerProcess.has_crashed):
68         (MockServerProcess.read_stdout_line):
69         (MockServerProcess.read_stdout):
70         (MockServerProcess.read_either_stdout_or_stderr_line):
71         (MockServerProcess.start):
72         (MockServerProcess.stop):
73         (MockServerProcess.kill):
74         * Scripts/webkitpy/layout_tests/port/gtk.py:
75         * Scripts/webkitpy/layout_tests/port/webkit.py:
76         (WebKitPort._read_image_diff):
77         (WebKitPort.skipped_layout_tests):
78         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
79         (test_path_to_apache_config_file):
80         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
81         (XvfbDriver):
82         (XvfbDriver.stop):
83
84 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
85
86         webkitpy.common.system.executive_unittest.ExecutiveTest.test_kill_process is flaky
87         https://bugs.webkit.org/show_bug.cgi?id=91698
88
89         Unreviewed, build fix.
90
91         This is a workaround for the test failing (apparently mostly?)
92         when run in parallel with other tests. The underlying bug still
93         needs to be diagnosed as we shouldn't be getting ECHILD from
94         process.wait().
95
96         * Scripts/webkitpy/common/system/executive_unittest.py:
97         (ExecutiveTest.test_kill_process):
98
99 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
100
101         Disable running python unit tests in parallel on win32 for now.
102         https://bugs.webkit.org/show_bug.cgi?id=91697
103
104         Unreviewed, build fix.
105
106         * Scripts/webkitpy/test/main.py:
107         (Tester._parse_args):
108
109 2012-07-18  Dave Tu  <dtu@chromium.org>
110
111         [chromium] Add gl_tests to flakiness dashboard.
112         https://bugs.webkit.org/show_bug.cgi?id=91680
113
114         Reviewed by Dirk Pranke.
115
116         * TestResultServer/static-dashboards/builders.js:
117         (loadBuildersList):
118         * TestResultServer/static-dashboards/dashboard_base.js:
119         (currentBuilderGroupCategory):
120
121 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
122
123         fix python unit tests failing under cygwin
124         https://bugs.webkit.org/show_bug.cgi?id=91678
125
126         Reviewed by Adam Barth.
127
128         Fix a case where executive.kill_process was failing under cygwin
129         (apple win port) in an unexpected way.
130         
131         * Scripts/webkitpy/common/system/executive.py:
132         (Executive.kill_process):
133
134 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
135
136         test-webkitpy: run tests in parallel by default
137         https://bugs.webkit.org/show_bug.cgi?id=91422
138
139         Reviewed by Adam Barth.
140
141         We use multiprocessing.cpu_count() for the default number of
142         jobs to run; memory overhead should be very low, so this should
143         be fine.
144
145         * Scripts/webkitpy/test/main.py:
146         (Tester._parse_args):
147
148 2012-07-18  Ryosuke Niwa  <rniwa@webkit.org>
149
150         Add Pravin D to the list of contributors.
151
152         * Scripts/webkitpy/common/config/committers.py:
153
154 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
155
156         [Qt] feature detection in orwt/nrwt does not work with force_static_libs_as_shared
157         https://bugs.webkit.org/show_bug.cgi?id=91514
158
159         Reviewed by Dirk Pranke.
160
161         Analyzing libQtWebKit.so is not enough in the case of force_static_libs_as_shared.
162         We need to analyze all the shared library or at least the WebCore one.
163
164         * Scripts/webkitdirs.pm:
165         (builtDylibPathForName):
166         In orwt it is hard coded to search for symbols in the WebCore library.
167         While in theory it is possible that symbols for a given feature are not
168         located in that, it doesn't happen in practice, so returning the path of
169         the WebCore library for a force_static_libs_as_shared build is enough to
170         fix the bug and it doesn't require a bigger refactoring. For a default
171         build we still return the path of the QtWebKit lib.
172         * Scripts/webkitpy/layout_tests/port/qt.py:
173         (QtPort._modules_to_search_for_symbols):
174         Enumerate all dynamic libraries, not just the QtWebKit one.
175
176 2012-07-18  Tony Chang  <tony@chromium.org>
177
178         [chromium] Unreviewed, more compile fixes on Chromium Win.
179
180         * DumpRenderTree/chromium/TestShellWin.cpp:
181         (TestShell::waitTestFinished):
182         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
183
184 2012-07-18  Christophe Dumez  <christophe.dumez@intel.com>
185
186         [EFL] Add central error management to EFL port
187         https://bugs.webkit.org/show_bug.cgi?id=91598
188
189         Reviewed by Kenneth Rohde Christiansen.
190
191         Map WebKitNetworkError to NSURLErrorDomain when
192         printing in DumpRenderTree so that the output
193         matches the expected one.
194
195         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
196         (descriptionSuitableForTestResult):
197
198 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
199
200         nrwt: start merging port/webkit.py into port/base.py
201         https://bugs.webkit.org/show_bug.cgi?id=91559
202
203         Reviewed by Ojan Vafai.
204
205         Since all the non-test port implementations now derive from
206         WebKitPort, there's no real point in keeping WebKitPort distinct
207         from Port. This patch starts merging the two by moving nearly
208         all of the webkit implementations of routines with no default
209         behavior into base.py. The few that didn't move rely on
210         additional infrastructure that should be refactored differently
211         (like the image diffing, which should probably be its own
212         class) and deserve their own patches.
213
214         This patch should just be moving code around, and require no
215         additional tests; in fact, we can delete the tests that were
216         asserting virtual methods in the base class.
217
218         * Scripts/webkitpy/layout_tests/port/base.py:
219         (Port.baseline_search_path):
220         (Port.check_build):
221         (Port):
222         (Port._check_driver):
223         (Port._check_port_build):
224         (Port.check_image_diff):
225         (Port.driver_name):
226         (Port.default_results_directory):
227         (Port.to.setup_environ_for_server):
228         (Port._path_to_apache):
229         (Port._is_redhat_based):
230         (Port._is_debian_based):
231         (Port._apache_config_file_name_for_platform):
232         (Port._path_to_apache_config_file):
233         (Port._build_path):
234         (Port._path_to_driver):
235         (Port._path_to_webcore_library):
236         (Port._path_to_helper):
237         (Port._path_to_image_diff):
238         (Port._path_to_wdiff):
239         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
240         (PortTest.test_httpd_returns_error_code):
241         * Scripts/webkitpy/layout_tests/port/webkit.py:
242         (WebKitPort):
243         (WebKitPort._build_driver_flags):
244         (WebKitPort._read_image_diff):
245         (WebKitPort.skipped_layout_tests):
246
247 2012-07-18  Tommy Widenflycht  <tommyw@google.com>
248
249         [chromium] MediaStream API: Moving the mock create* WebRTC calls into a shadow Platform class
250         https://bugs.webkit.org/show_bug.cgi?id=86215
251
252         Reviewed by Adam Barth.
253
254         Adding a shadow Platform object that is used to override some WebKit::Platform funtions to
255         instead create mock objects. No actual mock objects created yet.
256
257         * DumpRenderTree/DumpRenderTree.gypi:
258         * DumpRenderTree/chromium/DumpRenderTree.cpp:
259         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
260         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp: Copied from Tools/DumpRenderTree/chromium/config.h.
261         (MockWebKitPlatformSupport::create):
262         (MockWebKitPlatformSupport::MockWebKitPlatformSupport):
263         (MockWebKitPlatformSupport::cryptographicallyRandomValues):
264         (MockWebKitPlatformSupport::createMediaStreamCenter):
265         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h: Copied from Tools/DumpRenderTree/chromium/config.h.
266         (MockWebKitPlatformSupport):
267         * DumpRenderTree/chromium/config.h:
268
269 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
270
271         [CMake][EFL] Build and run TestWebKitAPI unit tests
272         https://bugs.webkit.org/show_bug.cgi?id=90671
273
274         Reviewed by Daniel Bates.
275
276         This patch enables almost for free a significant amount of test for the
277         WebKit C API, WTF and KURL. It makes easy for CMake based ports to
278         enable TestWebKitAPI since the missing bits to implement are
279         relatively simple.
280
281         * CMakeLists.txt:
282         * TestWebKitAPI/CMakeLists.txt: Added.
283         * TestWebKitAPI/PlatformEfl.cmake: Added.
284         * TestWebKitAPI/PlatformWebView.h:
285         * TestWebKitAPI/config.h:
286         * TestWebKitAPI/efl/InjectedBundleController.cpp: Added.
287         (TestWebKitAPI):
288         (TestWebKitAPI::InjectedBundleController::platformInitialize):
289         * TestWebKitAPI/efl/PlatformUtilities.cpp: Added.
290         (TestWebKitAPI):
291         (Util):
292         (TestWebKitAPI::Util::run):
293         (TestWebKitAPI::Util::sleep):
294         (TestWebKitAPI::Util::createURLForResource):
295         (TestWebKitAPI::Util::createInjectedBundlePath):
296         (TestWebKitAPI::Util::URLForNonExistentResource):
297         * TestWebKitAPI/efl/PlatformWebView.cpp: Added.
298         (TestWebKitAPI):
299         (TestWebKitAPI::initEcoreEvas):
300         (TestWebKitAPI::PlatformWebView::PlatformWebView):
301         (TestWebKitAPI::PlatformWebView::~PlatformWebView):
302         (TestWebKitAPI::PlatformWebView::page):
303         * TestWebKitAPI/efl/main.cpp: Added.
304         (checkForUseX11WindowArgument):
305         (main):
306
307 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
308
309         [CMake][EFL] Building jsc causes reconfiguration
310         https://bugs.webkit.org/show_bug.cgi?id=91387
311
312         Reviewed by Daniel Bates.
313
314         We should remove CMakeCache only when running build-webkit script,
315         otherwise it will cause a reconfiguration every time someone calls
316         generateBuildSystemFromCMakeProject(). We were re-building "jsc" and
317         not running WebKit2 unit tests on the bots because the project was
318         reconfigured with default values by the scripts that run these tests.
319
320         * Scripts/build-webkit:
321         * Scripts/webkitdirs.pm:
322         (removeCMakeCache):
323         (generateBuildSystemFromCMakeProject):
324
325 2012-07-18  Seokju Kwon  <seokju.kwon@samsung.com>
326
327         [EFL][DRT] Add support for Web Inspector in WebKit-EFL DRT
328         https://bugs.webkit.org/show_bug.cgi?id=87935
329
330         Reviewed by Andreas Kling.
331
332         Web Inspector will be shown when path or url contains "inspector/".
333         Dumprendertree should wait util web inspector resources are loaded totally
334         and handle the signals for creating or removing a view of web inspector.
335         ("inspector,view,create" and "inspector,view,close")
336
337         * DumpRenderTree/efl/DumpRenderTree.cpp:
338         (shouldOpenWebInspector):
339         (createLayoutTestController):
340         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
341         (DumpRenderTreeChrome::createView):
342         (DumpRenderTreeChrome::createWebInspectorView):
343         (DumpRenderTreeChrome::removeWebInspectorView):
344         (DumpRenderTreeChrome::waitInspectorLoadFinished):
345         (DumpRenderTreeChrome::onInspectorViewCreate):
346         (DumpRenderTreeChrome::onInspectorViewClose):
347         (DumpRenderTreeChrome::onInspectorFrameLoadFinished):
348         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
349         (DumpRenderTreeChrome):
350         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
351         (onConsoleMessage):
352         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
353         (LayoutTestController::showWebInspector):
354         (LayoutTestController::closeWebInspector):
355         (LayoutTestController::evaluateInWebInspector):
356
357 2012-07-18  Simon Hausmann  <simon.hausmann@nokia.com>
358
359         [ANGLE] On QT, use Bison and Flex during ANGLE build
360         https://bugs.webkit.org/show_bug.cgi?id=91108
361
362         Reviewed by Kenneth Rohde Christiansen.
363
364         * qmake/mkspecs/features/default_post.prf: Add support for variable_out to our generators, to allow
365         generating not only for SOURCES but also ANGLE_SOURCES (in this bug)
366
367 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
368
369         [Qt] plugin tests should not be disabled for WebKit1
370         https://bugs.webkit.org/show_bug.cgi?id=91604
371
372         Reviewed by Simon Hausmann.
373
374         Instead of not building TestNetscapePlugIn, we could programatically
375         disable actually loading it from WTR until https://bugs.webkit.org/show_bug.cgi?id=86620
376         has been solved, so we can still test plugins on WebKit1.
377
378         * Tools.pro:
379         * WebKitTestRunner/TestController.cpp:
380         (WTR::TestController::initialize):
381         * WebKitTestRunner/qt/TestControllerQt.cpp:
382         (WTR::TestController::initializeTestPluginDirectory):
383
384 2012-07-18  Mario Sanchez Prada  <msanchez@igalia.com>
385
386         [WK2][GTK] Implement AccessibilityUIElement in WKTR for GTK
387         https://bugs.webkit.org/show_bug.cgi?id=89223
388
389         Reviewed by Chris Fleizach.
390
391         Implemented AccessibilityUIElement in WKTR for GTK, based in the
392         implementation present in DumpRenderTree.
393
394         Added new files in gtk/ and updated the cross-platform ones.
395
396         * WebKitTestRunner/GNUmakefile.am:
397         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
398         (WTR):
399         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
400         (AccessibilityUIElement):
401         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityUIElementGtk.cpp: Added.
402         (WTR):
403         (WTR::attributeSetToString):
404         (WTR::checkElementState):
405         (WTR::indexRangeInTable):
406         (WTR::alterCurrentValue):
407         (WTR::AccessibilityUIElement::AccessibilityUIElement):
408         (WTR::AccessibilityUIElement::~AccessibilityUIElement):
409         (WTR::AccessibilityUIElement::isEqual):
410         (WTR::AccessibilityUIElement::getChildren):
411         (WTR::AccessibilityUIElement::getChildrenWithRange):
412         (WTR::AccessibilityUIElement::childrenCount):
413         (WTR::AccessibilityUIElement::elementAtPoint):
414         (WTR::AccessibilityUIElement::indexOfChild):
415         (WTR::AccessibilityUIElement::childAtIndex):
416         (WTR::AccessibilityUIElement::linkedUIElementAtIndex):
417         (WTR::AccessibilityUIElement::ariaOwnsElementAtIndex):
418         (WTR::AccessibilityUIElement::ariaFlowToElementAtIndex):
419         (WTR::AccessibilityUIElement::disclosedRowAtIndex):
420         (WTR::AccessibilityUIElement::rowAtIndex):
421         (WTR::AccessibilityUIElement::selectedChildAtIndex):
422         (WTR::AccessibilityUIElement::selectedChildrenCount):
423         (WTR::AccessibilityUIElement::selectedRowAtIndex):
424         (WTR::AccessibilityUIElement::titleUIElement):
425         (WTR::AccessibilityUIElement::parentElement):
426         (WTR::AccessibilityUIElement::disclosedByRow):
427         (WTR::AccessibilityUIElement::attributesOfLinkedUIElements):
428         (WTR::AccessibilityUIElement::attributesOfDocumentLinks):
429         (WTR::AccessibilityUIElement::attributesOfChildren):
430         (WTR::AccessibilityUIElement::allAttributes):
431         (WTR::AccessibilityUIElement::stringAttributeValue):
432         (WTR::AccessibilityUIElement::numberAttributeValue):
433         (WTR::AccessibilityUIElement::uiElementAttributeValue):
434         (WTR::AccessibilityUIElement::boolAttributeValue):
435         (WTR::AccessibilityUIElement::isAttributeSettable):
436         (WTR::AccessibilityUIElement::isAttributeSupported):
437         (WTR::AccessibilityUIElement::parameterizedAttributeNames):
438         (WTR::AccessibilityUIElement::role):
439         (WTR::AccessibilityUIElement::subrole):
440         (WTR::AccessibilityUIElement::roleDescription):
441         (WTR::AccessibilityUIElement::title):
442         (WTR::AccessibilityUIElement::description):
443         (WTR::AccessibilityUIElement::orientation):
444         (WTR::AccessibilityUIElement::stringValue):
445         (WTR::AccessibilityUIElement::language):
446         (WTR::AccessibilityUIElement::helpText):
447         (WTR::AccessibilityUIElement::x):
448         (WTR::AccessibilityUIElement::y):
449         (WTR::AccessibilityUIElement::width):
450         (WTR::AccessibilityUIElement::height):
451         (WTR::AccessibilityUIElement::clickPointX):
452         (WTR::AccessibilityUIElement::clickPointY):
453         (WTR::AccessibilityUIElement::intValue):
454         (WTR::AccessibilityUIElement::minValue):
455         (WTR::AccessibilityUIElement::maxValue):
456         (WTR::AccessibilityUIElement::valueDescription):
457         (WTR::AccessibilityUIElement::insertionPointLineNumber):
458         (WTR::AccessibilityUIElement::isActionSupported):
459         (WTR::AccessibilityUIElement::isEnabled):
460         (WTR::AccessibilityUIElement::isRequired):
461         (WTR::AccessibilityUIElement::isFocused):
462         (WTR::AccessibilityUIElement::isSelected):
463         (WTR::AccessibilityUIElement::isExpanded):
464         (WTR::AccessibilityUIElement::isChecked):
465         (WTR::AccessibilityUIElement::hierarchicalLevel):
466         (WTR::AccessibilityUIElement::speak):
467         (WTR::AccessibilityUIElement::ariaIsGrabbed):
468         (WTR::AccessibilityUIElement::ariaDropEffects):
469         (WTR::AccessibilityUIElement::lineForIndex):
470         (WTR::AccessibilityUIElement::rangeForLine):
471         (WTR::AccessibilityUIElement::rangeForPosition):
472         (WTR::AccessibilityUIElement::boundsForRange):
473         (WTR::AccessibilityUIElement::stringForRange):
474         (WTR::AccessibilityUIElement::attributedStringForRange):
475         (WTR::AccessibilityUIElement::attributedStringRangeIsMisspelled):
476         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
477         (WTR::AccessibilityUIElement::attributesOfColumnHeaders):
478         (WTR::AccessibilityUIElement::attributesOfRowHeaders):
479         (WTR::AccessibilityUIElement::attributesOfColumns):
480         (WTR::AccessibilityUIElement::attributesOfRows):
481         (WTR::AccessibilityUIElement::attributesOfVisibleCells):
482         (WTR::AccessibilityUIElement::attributesOfHeader):
483         (WTR::AccessibilityUIElement::rowCount):
484         (WTR::AccessibilityUIElement::columnCount):
485         (WTR::AccessibilityUIElement::indexInTable):
486         (WTR::AccessibilityUIElement::rowIndexRange):
487         (WTR::AccessibilityUIElement::columnIndexRange):
488         (WTR::AccessibilityUIElement::cellForColumnAndRow):
489         (WTR::AccessibilityUIElement::horizontalScrollbar):
490         (WTR::AccessibilityUIElement::verticalScrollbar):
491         (WTR::AccessibilityUIElement::selectedTextRange):
492         (WTR::AccessibilityUIElement::setSelectedTextRange):
493         (WTR::AccessibilityUIElement::increment):
494         (WTR::AccessibilityUIElement::decrement):
495         (WTR::AccessibilityUIElement::showMenu):
496         (WTR::AccessibilityUIElement::press):
497         (WTR::AccessibilityUIElement::setSelectedChild):
498         (WTR::AccessibilityUIElement::accessibilityValue):
499         (WTR::AccessibilityUIElement::documentEncoding):
500         (WTR::AccessibilityUIElement::documentURI):
501         (WTR::AccessibilityUIElement::url):
502         (WTR::AccessibilityUIElement::addNotificationListener):
503         (WTR::AccessibilityUIElement::removeNotificationListener):
504         (WTR::AccessibilityUIElement::isFocusable):
505         (WTR::AccessibilityUIElement::isSelectable):
506         (WTR::AccessibilityUIElement::isMultiSelectable):
507         (WTR::AccessibilityUIElement::isVisible):
508         (WTR::AccessibilityUIElement::isOffScreen):
509         (WTR::AccessibilityUIElement::isCollapsed):
510         (WTR::AccessibilityUIElement::isIgnored):
511         (WTR::AccessibilityUIElement::hasPopup):
512         (WTR::AccessibilityUIElement::takeFocus):
513         (WTR::AccessibilityUIElement::takeSelection):
514         (WTR::AccessibilityUIElement::addSelection):
515         (WTR::AccessibilityUIElement::removeSelection):
516         (WTR::AccessibilityUIElement::textMarkerRangeForElement):
517         (WTR::AccessibilityUIElement::textMarkerRangeLength):
518         (WTR::AccessibilityUIElement::previousTextMarker):
519         (WTR::AccessibilityUIElement::nextTextMarker):
520         (WTR::AccessibilityUIElement::stringForTextMarkerRange):
521         (WTR::AccessibilityUIElement::textMarkerRangeForMarkers):
522         (WTR::AccessibilityUIElement::startTextMarkerForTextMarkerRange):
523         (WTR::AccessibilityUIElement::endTextMarkerForTextMarkerRange):
524         (WTR::AccessibilityUIElement::textMarkerForPoint):
525         (WTR::AccessibilityUIElement::accessibilityElementForTextMarker):
526         (WTR::AccessibilityUIElement::attributedStringForTextMarkerRangeContainsAttribute):
527         (WTR::AccessibilityUIElement::indexForTextMarker):
528         (WTR::AccessibilityUIElement::isTextMarkerValid):
529         (WTR::AccessibilityUIElement::textMarkerForIndex):
530
531         Add documentEncoding and documentURI to AccessibilityUIElement.idl.
532
533         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
534
535 2012-07-18  Alexis Menard  <alexis.menard@openbossa.org>
536
537         [EFL] Build fix in WebKitTestRunner.
538         https://bugs.webkit.org/show_bug.cgi?id=91567
539
540         Reviewed by Kentaro Hara.
541
542         sleep() is defined in unistd.h, we need to include it.
543
544         * WebKitTestRunner/efl/TestControllerEfl.cpp:
545
546 2012-07-18  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
547
548         [NRWT] Unreviewed gardening after r122913
549         https://bugs.webkit.org/show_bug.cgi?id=91601
550
551         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
552         (TrivialMockPort.process_kill_time):
553
554 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
555
556         [Chromium-Android] Run ref tests together to avoid expensive driver restarts
557         https://bugs.webkit.org/show_bug.cgi?id=91533
558
559         Reviewed by Dirk Pranke.
560
561         Though DriverProxy maintains two drivers to support pixel tests and non-pixel tests,
562         chromium-android uses another way because it can't support multiple drivers.
563         It restarts the driver when pixel-test mode changes (e.g. when running a ref test after
564         a normal test in --no-pixel-tests mode). However restarting driver is expensive on
565         Android (several seconds each time). To reduce the cost, a command line option
566         '--shard-ref-tests' is added to group ref tests in dedicated shards.
567         The option is by default enabled on Android.
568
569         Will remove the option once DRT supports switching pixel test mode during one run.
570         (https://bugs.webkit.org/show_bug.cgi?id=91538, https://bugs.webkit.org/show_bug.cgi?id=91539)
571
572         * Scripts/webkitpy/layout_tests/controllers/manager.py:
573         (Manager._shard_tests):
574         (Manager._shard_in_two):
575         (Manager._shard_by_directory):
576         (Manager._run_tests):
577         * Scripts/webkitpy/layout_tests/controllers/worker.py:
578         (Worker._update_test_input):
579         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
580         (ChromiumAndroidPort.__init__):
581         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
582         (parse_args):
583
584 2012-07-17  Don Olmstead  <don.olmstead@am.sony.com>
585
586         NRWT The time before server_process kills DRT should be variable
587         https://bugs.webkit.org/show_bug.cgi?id=91542
588
589         Reviewed by Dirk Pranke.
590
591         Added the ability to configure the amount of time
592         server_process waits before killing the DRT process
593         during the call to stop.
594
595         * Scripts/webkitpy/layout_tests/port/base.py:
596         (Port.variable.process_kill_time):
597         * Scripts/webkitpy/layout_tests/port/server_process.py:
598         (ServerProcess.stop):
599
600 2012-07-17  Alexis Menard  <alexis.menard@openbossa.org>
601
602         webkitdirs.pm should fallback to uname -m if arch is not present.
603         https://bugs.webkit.org/show_bug.cgi?id=91543
604
605         Reviewed by Martin Robinson.
606
607         It seems that coreutils is moving away from the 'arch' command. Archlinux
608         for example doesn't ship it anymore (coreutils 8.17). We can then fallback
609         to 'uname -m' to find out the architecture.
610
611         * Scripts/webkitdirs.pm:
612         (determineArchitecture):
613
614 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
615
616         nrwt: rename printer.print_update to printer.write_update to match metered_stream
617         https://bugs.webkit.org/show_bug.cgi?id=91557
618
619         Reviewed by Ojan Vafai.
620
621         Now that all of the printing logic is in printer.py, I want the
622         public interface to printer to follow the meteredstream
623         interface where possible. renaming write_update() gets us close.
624
625         * Scripts/webkitpy/layout_tests/controllers/manager.py:
626         (Manager._run_tests):
627         (Manager._set_up_run):
628         (Manager.start_servers_with_lock):
629         (Manager.stop_servers_with_lock):
630         (Manager._clobber_old_results):
631         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
632         (run):
633         * Scripts/webkitpy/layout_tests/views/printing.py:
634         (Printer.write_update):
635         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
636         (Testprinter.test_write_update):
637
638 2012-07-17  Tony Chang  <tony@chromium.org>
639
640         Fix a bug where optimize-baselines would incorrectly fail to optimize
641         https://bugs.webkit.org/show_bug.cgi?id=91551
642
643         Reviewed by Adam Barth.
644
645         In some cases, optimize-baseline would correctly optimize the results, but because
646         we weren't filtering the virtual ports out of _results_by_port_name, we thought
647         that we had failed to optimize.
648
649         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
650         (BaselineOptimizer._find_optimal_result_placement): No virtual filtering here.
651         (BaselineOptimizer._filtered_results_by_port_name): New function that filters out virtual directories.
652         (BaselineOptimizer.optimize): Filter out virtual ports.
653         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
654         (TestBaselineOptimizer._move_baselines): Add a stub so we don't actual move results.
655         (BaselineOptimizerTest._assertOptimization): Call optimize and verify that the right files were moved.
656         (BaselineOptimizerTest._assertOptimizationFailed): Add a method for when optimization should fail.
657         (BaselineOptimizerTest.test_common_directory_includes_root): Update since this test should fail.
658         (BaselineOptimizerTest.test_virtual_ports_filtered): New test case that demonstrates the bug.
659
660 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
661
662         [Chromium] Add --encode-binary command line option for DRT
663         https://bugs.webkit.org/show_bug.cgi?id=91532
664
665         Reviewed by Adam Barth.
666
667         When the option presents, DRT will encode binary output data in base64.
668         ChromiumAndroidPort will pass the option because 'adb shell' doesn't support
669         binary data output.
670
671         This change is a refactoring to replace the original hard-coded code for Android.
672
673         * DumpRenderTree/chromium/DumpRenderTree.cpp:
674         (main):
675         * DumpRenderTree/chromium/TestEventPrinter.cpp:
676         (TestEventPrinter::TestEventPrinter):
677         (TestEventPrinter::handleAudio): New function replacing the original handleAudioHeader
678         (TestEventPrinter::handleImage):
679         (TestEventPrinter::handleBinary): Extracted common function for handleAudio and handleImage
680         * DumpRenderTree/chromium/TestEventPrinter.h:
681         (TestEventPrinter):
682         * DumpRenderTree/chromium/TestShell.cpp:
683         (TestShell::TestShell):
684         (TestShell::initialize):
685         (TestShell::dump):
686         * DumpRenderTree/chromium/TestShell.h:
687         (TestShell):
688         (TestShell::setEncodeBinary):
689         * Scripts/webkitpy/layout_tests/port/chromium_android.py: Add --encode-binary to additional_drt_flags
690         (ChromiumAndroidPort.__init__):
691
692 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
693
694         nrwt: move per-test result output into printing.py
695         https://bugs.webkit.org/show_bug.cgi?id=91465
696
697         Reviewed by Ojan Vafai.
698
699         This moves the last substantive logging/printing code from the
700         manager into printing.py. Subsequent patches in this thread can
701         largely focus on cleaning up printing.py and will leave the rest
702         of the code alone.
703
704         This patch also removes manager.update() and
705         manager.update_summary(), two routines that were no longer being
706         used or needed (they have been dead for a long time but I didn't
707         notice them until now).
708
709         * Scripts/webkitpy/layout_tests/controllers/manager.py:
710         (Manager.results_directory):
711         (Manager._look_for_new_crash_logs):
712         (Manager._update_summary_with_result):
713         * Scripts/webkitpy/layout_tests/views/printing.py:
714         (Printer.print_finished_test):
715
716 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
717
718         nrwt: move config-specific logging to printing.py
719         https://bugs.webkit.org/show_bug.cgi?id=91450
720
721         Reviewed by Ojan Vafai.
722
723         More refactoring ... this moves the 'config' output to
724         printing.py.
725
726         * Scripts/webkitpy/layout_tests/controllers/manager.py:
727         (Manager._resize_shards):
728         (Manager._run_tests):
729         (Manager._upload_json_files):
730         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
731         (ManagerTest.get_options):
732         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
733         (run):
734         * Scripts/webkitpy/layout_tests/views/printing.py:
735         (Printer.print_config):
736         (Printer.print_workers_and_shards):
737         (Printer._print_config):
738         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
739         (Testprinter.test_fallback_path_in_config):
740         (Testprinter.test_print_config):
741
742 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
743
744         nrwt: clean up logging when we're only running one chunk of the tests
745         https://bugs.webkit.org/show_bug.cgi?id=91447
746
747         Reviewed by Ojan Vafai.
748
749         Apparently we were creating a 'tests_run.txt' in the results
750         directory if we were only running one chunk of the tests; after
751         3+ years of hacking on this code, that was news to me, so I'm
752         guessing no one uses this. I'm removing this, and changing the
753         _print_expected calls to debug messages to simplify the
754         layering.
755
756         * Scripts/webkitpy/layout_tests/controllers/manager.py:
757         (Manager._split_into_chunks_if_necessary):
758
759 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
760
761         Fix regression in style checker introduced in r122868.
762         https://bugs.webkit.org/show_bug.cgi?id=91470
763
764         Unreviewed, build fix.
765
766         * Scripts/webkitpy/style/checkers/cpp.py:
767         (_FileState.__init__):
768
769 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
770
771         nrwt: move the bulk of the "expected" output to printing.py
772         https://bugs.webkit.org/show_bug.cgi?id=91442
773
774         Reviewed by Ojan Vafai.
775
776         More printing-related refactoring. This moves all of the code
777         that prints the results we expect to get, but doesn't move a few
778         dangling printfs (those'll get hit in a later patch).
779         
780         No functional changes; covered by existing tests.
781
782         * Scripts/webkitpy/layout_tests/controllers/manager.py:
783         (Manager._split_into_chunks_if_necessary):
784         (Manager.prepare_lists_and_print_output):
785         (Manager.print_config):
786         * Scripts/webkitpy/layout_tests/views/printing.py:
787         (Printer.print_expected):
788         (Printer):
789         (Printer._print_expected_results_of_type):
790         (Printer._num_digits):
791         (Printer._print_expected):
792         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
793         (Testprinter.test_print_expected):
794
795 2012-07-17  Christophe Dumez  <christophe.dumez@intel.com>
796
797         [EFL] Replace 0 by NULL in public headers documentation
798         https://bugs.webkit.org/show_bug.cgi?id=91470
799
800         Reviewed by Dirk Pranke.
801
802         Fix style checking to properly detect C headers as C
803         files. Without this change, the style script would
804         complain if we use NULL (instead of 0 / null) in C
805         headers.
806
807         * Scripts/webkitpy/style/checkers/cpp.py:
808         (_FileState.__init__):
809         (_FileState.is_c):
810         (_FileState.is_c_or_objective_c):
811
812 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
813
814         [chromium] remove --test-shell support from DRT
815         https://bugs.webkit.org/show_bug.cgi?id=86927
816
817         Reviewed by Adam Barth.
818
819         Now no one use --test-shell. All of us use the standard DRT mode.
820
821         * DumpRenderTree/chromium/DumpRenderTree.cpp: Removed support of --test-shell and --pixel-tests=filename options. (--pixel-tests without '=' is kept for DRT mode.)
822         (runTest):
823         (main):
824         * DumpRenderTree/chromium/TestEventPrinter.cpp:
825         (TestEventPrinter::TestEventPrinter):
826         (TestEventPrinter::~TestEventPrinter):
827         (TestEventPrinter::handleTestHeader):
828         (TestEventPrinter::handleTimedOut):
829         (TestEventPrinter::handleTextHeader):
830         (TestEventPrinter::handleTextFooter):
831         (TestEventPrinter::handleAudioHeader):
832         (TestEventPrinter::handleAudioFooter):
833         (TestEventPrinter::handleImage): Removed the unused filename parameter.
834         (TestEventPrinter::handleTestFooter):
835         * DumpRenderTree/chromium/TestEventPrinter.h:
836         (TestEventPrinter): Made it a concrete class because we have only one implementation.
837         * DumpRenderTree/chromium/TestShell.cpp:
838         (TestShell::TestShell):
839         (TestShell::initialize):
840         (TestShell::dumpImage):
841         * DumpRenderTree/chromium/TestShell.h:
842         (TestParams): Removed pixelFileName field.
843         (TestShell):
844
845 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
846
847         remove ChromiumDriver from NRWT
848         https://bugs.webkit.org/show_bug.cgi?id=88478
849
850         Now WebKitDriver has replaced ChromiumDriver since test_shell mode is deperecated.
851
852         Reviewed by Dirk Pranke.
853
854         * Scripts/webkitpy/layout_tests/port/chromium.py: Removed ChromiumDriver code.
855         (ChromiumPort._driver_class): Removed. WebKitPort._driver_class() will be used.
856         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
857
858 2012-07-17  No'am Rosenthal  <noam.rosenthal@nokia.com>
859
860         Unreviewed. Add QtGraphics, TextureMapper and OpenGL to watchlist.
861
862         * Scripts/webkitpy/common/config/watchlist:
863
864 2012-07-17  Sheriff Bot  <webkit.review.bot@gmail.com>
865
866         Unreviewed, rolling out r122828.
867         http://trac.webkit.org/changeset/122828
868         https://bugs.webkit.org/show_bug.cgi?id=91516
869
870         DumpRenderTree crashes after printing test results. (Requested
871         by vsevik on #webkit).
872
873         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
874         (TestInterfaces::TestInterfaces):
875         (TestInterfaces::~TestInterfaces):
876         (TestInterfaces::bindTo):
877         (TestInterfaces::resetAll):
878         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
879         (TestInterfaces):
880         * DumpRenderTree/chromium/TestShell.cpp:
881         (TestShell::initialize):
882         (TestShell::createMainWindow):
883         (TestShell::~TestShell):
884         (TestShell::resetTestController):
885         (TestShell::bindJSObjectsToWindow):
886         * DumpRenderTree/chromium/TestShell.h:
887         (TestShell::accessibilityController):
888         (TestShell):
889
890 2012-07-17  Gabor Ballabas  <gaborb@inf.u-szeged.hu>
891
892         [Qt][V8] Remove the V8 related codepaths and configuration
893         https://bugs.webkit.org/show_bug.cgi?id=90863
894
895         Reviewed by Simon Hausmann.
896
897         * Scripts/build-webkit: Remove --v8 option.
898
899 2012-07-17  Sheriff Bot  <webkit.review.bot@gmail.com>
900
901         Unreviewed, rolling out r122834.
902         http://trac.webkit.org/changeset/122834
903         https://bugs.webkit.org/show_bug.cgi?id=91492
904
905         it broke the chromium (Requested by kkristof on #webkit).
906
907         * Scripts/build-webkit:
908
909 2012-07-17  Gabor Ballabas  <gaborb@inf.u-szeged.hu>
910
911         [Qt][V8] Remove the V8 related codepaths and configuration
912         https://bugs.webkit.org/show_bug.cgi?id=90863
913
914         Reviewed by Simon Hausmann.
915
916         * Scripts/build-webkit: Remove --v8 option.
917
918 2012-07-17  Balazs Kelemen  <kbalazs@webkit.org>
919
920         [Qt] Add --use-test-fonts option to Minibrowser
921         https://bugs.webkit.org/show_bug.cgi?id=91402
922
923         Reviewed by Tor Arne Vestbø.
924
925         Same stuff that already exists for QtTestBrowser.
926
927         * MiniBrowser/qt/MiniBrowser.pro:
928         * MiniBrowser/qt/MiniBrowserApplication.cpp:
929         (MiniBrowserApplication::handleUserOptions):
930
931 2012-07-17  Adam Barth  <abarth@webkit.org>
932
933         [Chromium] TestInterfaces should be responsible for owning and binding AccessibilityController and TextInputController
934         https://bugs.webkit.org/show_bug.cgi?id=91459
935
936         Reviewed by Ryosuke Niwa.
937
938         Rather than having TestShell own and bind each of these JavaScript APIs
939         individually, TestShell should delegate that work to TestInterfaces.
940         This patch moves AccessibilityController and TextInputController to be
941         the responsibility of TestInterfaces.
942
943         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
944         (TestInterfaces::TestInterfaces):
945         (TestInterfaces::~TestInterfaces):
946         (TestInterfaces::setWebView):
947         (TestInterfaces::bindTo):
948         (TestInterfaces::resetAll):
949         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
950         (WebKit):
951         (TestInterfaces):
952         (TestInterfaces::accessibilityController):
953         * DumpRenderTree/chromium/TestShell.cpp:
954         (TestShell::initialize):
955         (TestShell::createMainWindow):
956         (TestShell::~TestShell):
957         (TestShell::resetTestController):
958         (TestShell::bindJSObjectsToWindow):
959         * DumpRenderTree/chromium/TestShell.h:
960         (TestShell::accessibilityController):
961         (TestShell):
962
963 2012-07-17  Adam Barth  <abarth@webkit.org>
964
965         [Chromium] Move TextInputController into TestRunner.a
966         https://bugs.webkit.org/show_bug.cgi?id=91457
967
968         Reviewed by Ryosuke Niwa.
969
970         This patch is purely moving code. TextInputController is one of the
971         objects exposed to JavaScript and so should be in the TestRunner
972         library.
973
974         * DumpRenderTree/DumpRenderTree.gypi:
975         * DumpRenderTree/chromium/TestRunner/TextInputController.cpp: Renamed from Tools/DumpRenderTree/chromium/TextInputController.cpp.
976         * DumpRenderTree/chromium/TestRunner/TextInputController.h: Renamed from Tools/DumpRenderTree/chromium/TextInputController.h.
977
978 2012-07-17  Adam Barth  <abarth@webkit.org>
979
980         [Chromium] TextInputController shouldn't know about TestShell
981         https://bugs.webkit.org/show_bug.cgi?id=91455
982
983         Reviewed by Kent Tamura.
984
985         This patch uses the same pattern we used for the
986         AccessibilityController to remove the dependency on TestShell.
987
988         * DumpRenderTree/chromium/TestShell.cpp:
989         (TestShell::initialize):
990         (TestShell::createMainWindow):
991         (TestShell::~TestShell):
992         * DumpRenderTree/chromium/TextInputController.cpp:
993         (TextInputController::TextInputController):
994         (TextInputController::insertText):
995         (TextInputController::doCommand):
996         (TextInputController::setMarkedText):
997         (TextInputController::unmarkText):
998         (TextInputController::hasMarkedText):
999         (TextInputController::markedRange):
1000         (TextInputController::selectedRange):
1001         (TextInputController::firstRectForCharacterRange):
1002         (TextInputController::validAttributesForMarkedText):
1003         (TextInputController::setComposition):
1004         * DumpRenderTree/chromium/TextInputController.h:
1005         (WebKit):
1006         (TextInputController):
1007         (TextInputController::setWebView):
1008
1009 2012-07-17  David Barr  <davidbarr@chromium.org>
1010
1011         Introduce ENABLE_CSS_IMAGE_ORIENTATION compile flag
1012         https://bugs.webkit.org/show_bug.cgi?id=89055
1013
1014         Reviewed by Kent Tamura.
1015
1016         The css3-images module is at candidate recommendation.
1017         http://www.w3.org/TR/2012/CR-css3-images-20120417/#the-image-orientation
1018
1019         Add a configuration option for CSS image-orientation support, disabling it by default.
1020
1021         * Scripts/webkitperl/FeatureList.pm:
1022         * qmake/mkspecs/features/features.pri:
1023
1024 2012-07-16  Adam Barth  <abarth@webkit.org>
1025
1026         [Chromium] Move Task.(cpp|h) into TestRunner.a
1027         https://bugs.webkit.org/show_bug.cgi?id=91446
1028
1029         Reviewed by Ryosuke Niwa.
1030
1031         Task is a dependency of EventSender, which needs to move into TestRunner.a.
1032
1033         * DumpRenderTree/DumpRenderTree.gypi:
1034         * DumpRenderTree/chromium/TestRunner/Task.cpp: Renamed from Tools/DumpRenderTree/chromium/Task.cpp.
1035         * DumpRenderTree/chromium/TestRunner/Task.h: Renamed from Tools/DumpRenderTree/chromium/Task.h.
1036
1037 2012-07-16  Adam Barth  <abarth@webkit.org>
1038
1039         [Chromium] Move AccessibilityController and AccessibilityUIElement into TestRunner.a
1040         https://bugs.webkit.org/show_bug.cgi?id=91443
1041
1042         Reviewed by Tony Chang.
1043
1044         Purely a file move (and a sorting of #includes).
1045
1046         * DumpRenderTree/DumpRenderTree.gypi:
1047         * DumpRenderTree/chromium/TestRunner/AccessibilityController.cpp: Renamed from Tools/DumpRenderTree/chromium/AccessibilityController.cpp.
1048         * DumpRenderTree/chromium/TestRunner/AccessibilityController.h: Renamed from Tools/DumpRenderTree/chromium/AccessibilityController.h.
1049         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElement.cpp: Renamed from Tools/DumpRenderTree/chromium/AccessibilityUIElement.cpp.
1050         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElement.h: Renamed from Tools/DumpRenderTree/chromium/AccessibilityUIElement.h.
1051
1052 2012-07-16  Adam Barth  <abarth@webkit.org>
1053
1054         [Chromium] AccessibilityController shouldn't know anything about TestShell
1055         https://bugs.webkit.org/show_bug.cgi?id=91441
1056
1057         Reviewed by Tony Chang.
1058
1059         This patch is a precursor to moving AccessibilityController.cpp into TestRunner.a.
1060
1061         * DumpRenderTree/chromium/AccessibilityController.cpp:
1062         (AccessibilityController::AccessibilityController):
1063         (AccessibilityController::getFocusedElement):
1064         (AccessibilityController::getRootElement):
1065         * DumpRenderTree/chromium/AccessibilityController.h:
1066         (WebKit):
1067         (AccessibilityController):
1068         (AccessibilityController::setWebView):
1069         * DumpRenderTree/chromium/TestShell.cpp:
1070         (TestShell::initialize):
1071         (TestShell::createMainWindow):
1072         (TestShell::~TestShell):
1073
1074 2012-07-16  Adam Barth  <abarth@webkit.org>
1075
1076         [Chromium] Introduce TestInterfaces to hold all the JavaScript interfaces needed for LayoutTests
1077         https://bugs.webkit.org/show_bug.cgi?id=91312
1078
1079         Reviewed by Ryosuke Niwa.
1080
1081         Looking forward to moving more objects into TestRunner.a, we're going
1082         to need an object to own all the interfaces and to put them through
1083         their lifecycle.
1084
1085         * DumpRenderTree/DumpRenderTree.gypi:
1086         * DumpRenderTree/chromium/TestShell.cpp:
1087         (TestShell::initialize):
1088         (TestShell::resetTestController):
1089         (TestShell::bindJSObjectsToWindow):
1090         * DumpRenderTree/chromium/TestShell.h:
1091         (TestShell):
1092         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp: Added.
1093         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h: Added.
1094
1095 2012-07-16  Ryuan Choi  <ryuan.choi@samsung.com>
1096
1097         [EFL][WK2] Add APIs to support theme.
1098         https://bugs.webkit.org/show_bug.cgi?id=90107
1099
1100         Reviewed by Hajime Morita.
1101
1102         Override default theme path for MiniBrowser and WebKitTestRunner.
1103
1104         * MiniBrowser/efl/CMakeLists.txt:
1105         * MiniBrowser/efl/main.c:
1106         (browserCreate):
1107         * WebKitTestRunner/PlatformEfl.cmake:
1108         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1109         (WTR::PlatformWebView::PlatformWebView):
1110
1111 2012-07-16  Ryuan Choi  <ryuan.choi@samsung.com>
1112
1113         [EFL][DRT] Implement dumpFrameScrollPosition
1114         https://bugs.webkit.org/show_bug.cgi?id=87638
1115
1116         Reviewed by Hajime Morita.
1117
1118         * DumpRenderTree/efl/DumpRenderTree.cpp:
1119         (dumpFrameScrollPosition): Implemented.
1120         (shouldDumpFrameScrollPosition): Fixed wrong condition like other ports.
1121
1122 2012-07-16  Don Olmstead  <don.olmstead@am.sony.com>
1123
1124         NRWRT Should provide a VS project to work on
1125         https://bugs.webkit.org/show_bug.cgi?id=91436
1126
1127         Reviewed by Dirk Pranke.
1128
1129         Adding VS2010 project for modifying webkitpy. Uses pytools
1130         <http://pytools.codeplex.com> which integrates python into 
1131         Visual Studio.
1132
1133         * Scripts/webkitpy/webkitpy.pyproj: Added.
1134         * Scripts/webkitpy/webkitpy.sln: Added.
1135
1136 2012-07-16  Adam Barth  <abarth@webkit.org>
1137
1138         [Chromium] Move GamepadController into TestRunner.a
1139         https://bugs.webkit.org/show_bug.cgi?id=91311
1140
1141         Reviewed by Tony Chang.
1142
1143         GamepadController looks like a simple class to move into TestRunner.a
1144         because it has almost zero dependence on TestShell.
1145
1146         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1147         * DumpRenderTree/DumpRenderTree.gypi:
1148         * DumpRenderTree/chromium/TestRunner/CppBoundClass.cpp: Renamed from Tools/DumpRenderTree/chromium/CppBoundClass.cpp.
1149         * DumpRenderTree/chromium/TestRunner/CppBoundClass.h: Renamed from Tools/DumpRenderTree/chromium/CppBoundClass.h.
1150         * DumpRenderTree/chromium/TestRunner/CppVariant.cpp: Renamed from Tools/DumpRenderTree/chromium/CppVariant.cpp.
1151         * DumpRenderTree/chromium/TestRunner/CppVariant.h: Renamed from Tools/DumpRenderTree/chromium/CppVariant.h.
1152             - GamepadController depends on these bindings helpers.
1153         * DumpRenderTree/chromium/TestRunner/GamepadController.cpp: Renamed from Tools/DumpRenderTree/chromium/GamepadController.cpp.
1154         * DumpRenderTree/chromium/TestRunner/GamepadController.h: Renamed from Tools/DumpRenderTree/chromium/GamepadController.h.
1155             - Just moved these files and removed the unused TestShell
1156               references.
1157         * DumpRenderTree/chromium/TestRunner/Stub.cpp: Removed.
1158             - No longer needed.
1159         * DumpRenderTree/chromium/TestShell.cpp:
1160             - Update call to constructor to avoid passing in this.
1161
1162 2012-07-16  Xianzhu Wang  <wangxianzhu@chromium.org>
1163
1164         [Chromium-android] Don't use test_shell mode of DRT
1165         https://bugs.webkit.org/show_bug.cgi?id=88542
1166
1167         Reviewed by Dirk Pranke.
1168
1169         Test shell mode is about to be removed.
1170         Switch to use DRT mode for chromium-android.
1171
1172         Summary of changes:
1173         1. ChromiumAndroidDriver now inherits from WebKitDriver instead of ChromiumDriver (to be deprecated).
1174         2. Conforms to the DRT mode protocol for input/output of DumpRenderTree.
1175         3. Added support for Android 'adb shell' input/output (base64, newline mode, etc.)
1176
1177         * DumpRenderTree/chromium/TestEventPrinter.cpp:
1178         (DRTPrinter::handleImage): Outputs base64 on Android.
1179         * DumpRenderTree/chromium/TestShellAndroid.cpp:
1180         (platformInit): Changed err_file to err_fifo, required by python ServerProcess.
1181         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1182         (ChromiumAndroidPort.__init__):
1183         (ChromiumAndroidPort.create_driver): Override to create driver without DriverProxy to ensure 1 Driver per run.
1184         (ChromiumAndroidDriver):
1185         (ChromiumAndroidDriver.__init__):
1186         (ChromiumAndroidDriver.cmd_line):
1187         (ChromiumAndroidDriver._deadlock_detector):
1188         (ChromiumAndroidDriver._drt_cmd_line):
1189         (ChromiumAndroidDriver.start):
1190         (ChromiumAndroidDriver._start):
1191         (ChromiumAndroidDriver._start_once):
1192         (ChromiumAndroidDriver.run_test):
1193         (ChromiumAndroidDriver.stop):
1194         (ChromiumAndroidDriver._command_from_driver_input):
1195         (ChromiumAndroidDriver._read_prompt):
1196         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1197         (ChromiumAndroidPortTest.test_driver_cmd_line):
1198         (ChromiumAndroidDriverTest.test_cmd_line):
1199         (ChromiumAndroidDriverTest):
1200         (ChromiumAndroidDriverTest.test_drt_cmd_line):
1201         (ChromiumAndroidDriverTest.test_read_prompt):
1202         (ChromiumAndroidDriverTest.test_command_from_driver_input):
1203         (ChromiumAndroidDriverTest.test_write_command_and_read_line):
1204         * Scripts/webkitpy/layout_tests/port/server_process.py:
1205         (ServerProcess.__init__): Added universal_newlines to handle Android 'adb shell' line ends.
1206         (ServerProcess._start):
1207         (ServerProcess._wait_for_data_and_update_buffers_using_select): Handles unexpected EOF which indicates crash on Android.
1208         (ServerProcess.stop): Added kill_directly parameter to kill the process without waiting it (which always timeouts for Android).
1209         (ServerProcess.replace_outputs): Added to combine different input/output pipes into one ServerProcess.
1210         * Scripts/webkitpy/layout_tests/port/webkit.py:
1211         (WebKitDriver._command_from_driver_input):
1212         (WebKitDriver.run_test): Changed timeout origin so that slow start() on Android won't cause timeout of layout test case.
1213         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1214         (MockServerProcess.read_stdout):
1215         (MockServerProcess.start):
1216         (MockServerProcess):
1217         (MockServerProcess.stop):
1218         (MockServerProcess.kill):
1219         (WebKitDriverTest.test_read_block):
1220         (WebKitDriverTest.test_read_binary_block):
1221         (WebKitDriverTest.test_read_base64_block):
1222
1223 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1224
1225         nrwt: move a bunch of printing code from manager.py to printing.py
1226         https://bugs.webkit.org/show_bug.cgi?id=91439
1227
1228         Reviewed by Ojan Vafai.
1229
1230         All of the logic that handles what gets logged should eventually
1231         live in printing.py; this patch moves a large chunk of code that
1232         prints all of the stuff after the run completes.
1233
1234         There are no functional changes, this is just moving code
1235         around. Covered by existing tests (updated as necessary).
1236
1237         This code is all pretty messy and most of it should be deleted
1238         (or extracted after the fact from results.json), but I'm saving
1239         that for further patches. My goal is just to manager.py to a
1240         more manageable state.
1241
1242         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1243         (Manager.prepare_lists_and_print_output):
1244         (Manager.run):
1245         (Manager._mark_interrupted_tests_as_skipped):
1246         (Manager._update_summary_with_result):
1247         (Manager._num_digits):
1248         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1249         (ManagerTest.test_interrupt_if_at_failure_limits):
1250         (ResultSummaryTest.summarized_results):
1251         * Scripts/webkitpy/layout_tests/models/result_summary.py:
1252         (ResultSummary.__init__):
1253         (ResultSummary.add):
1254         Here we had to add a list of the slow tests to the result
1255         summary so that we didn't need to call back into the manager and
1256         the test expectations object to figure out if a test is SLOW.
1257         * Scripts/webkitpy/layout_tests/views/printing.py:
1258         (Printer.print_results):
1259         (Printer):
1260         (Printer._print_timing_statistics):
1261         (Printer._print_aggregate_test_statistics):
1262         (Printer._print_individual_test_times):
1263         (Printer._print_test_list_timing):
1264         (Printer._print_directory_timings):
1265         (Printer._print_statistics_for_test_timings):
1266         (Printer._print_result_summary):
1267         (Printer._print_result_summary_entry):
1268         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1269         (Testprinter.test_print_unexpected_results.get_unexpected_results):
1270         (Testprinter):
1271
1272 2012-07-16  Adam Barth  <abarth@webkit.org>
1273
1274         [Chromium] Create a stub TestRunner.a target
1275         https://bugs.webkit.org/show_bug.cgi?id=91309
1276
1277         Reviewed by Ryosuke Niwa.
1278
1279         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1280         * DumpRenderTree/DumpRenderTree.gypi:
1281         * DumpRenderTree/chromium/TestRunner/Stub.cpp: Added.
1282
1283 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1284
1285         test-webkitpy: clean up logging to make it ninja-esque
1286         https://bugs.webkit.org/show_bug.cgi?id=91297
1287
1288         Reviewed by Adam Barth.
1289
1290         This patch changes the output of test-webkitpy to be closer to
1291         what ninja produces. Namely:
1292
1293         If you are running with stderr writing to a tty (and not -v), we
1294         will no longer print '...' as tests complete. Instead, we use a
1295         metered stream and print entries of the form:
1296            '[X/Y] test_name passed/failed/erred'
1297         where X is the current test # and Y is the total number of tests.
1298
1299         If you are running with stderr piped to a file or other non-tty
1300         object (or with -v), you get the same output, one per line for
1301         every test.
1302
1303         In addition, if tests fail or err out, you get the stack trace
1304         immediately; you don't have to wait until the end of the run.
1305
1306         Lastly, this change cleans up the unit tests for test-webkitpy
1307         itself to not confuse the logger and to work cleanly w/ multiple
1308         processes. It looks like between this and all the clean up in
1309         MessagePool that happened as it landed, we can now run
1310         multiprocessing tests in parallel.
1311
1312         * Scripts/webkitpy/test/finder.py:
1313         (Finder._default_names):
1314         * Scripts/webkitpy/test/finder_unittest.py:
1315         (FinderTest.setUp):
1316         (FinderTest.tearDown):
1317         * Scripts/webkitpy/test/printer.py:
1318         (Printer.__init__):
1319         (Printer.configure):
1320         (Printer.print_started_test):
1321         (Printer.print_finished_test):
1322         (Printer._test_line):
1323         (Printer.print_result):
1324         * Scripts/webkitpy/test/runner.py:
1325         (Runner.run):
1326         * Scripts/webkitpy/test/runner_unittest.py:
1327         (RunnerTest.setUp):
1328         (RunnerTest):
1329         (RunnerTest.tearDown):
1330         (RunnerTest.assert_run):
1331         (RunnerTest.test_regular):
1332         (RunnerTest.test_verbose):
1333         (RunnerTest.test_timing):
1334
1335 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1336
1337         test-webkitpy: handle failures properly when running in parallel
1338         https://bugs.webkit.org/show_bug.cgi?id=91416
1339
1340         Reviewed by Adam Barth.
1341
1342         It turns out that unittest.TestResults contain a handle to the
1343         test method itself, which isn't picklable; it's sufficient to just
1344         store the test name instead of the actual method. By doing so
1345         we can move the test_name() method from the printer to the
1346         runner where it belongs (so the printer is less dependent on the
1347         unittest framework's data structures).
1348
1349         This change should really have a test but I don't know how to
1350         write one that properly captures the behavior and won't cause
1351         test-webkitpy itself to fail. I've verified the fix by hand, at
1352         least, in the meantime.
1353
1354         * Scripts/webkitpy/test/printer.py:
1355         (Printer.__init__):
1356         (Printer.print_result):
1357         * Scripts/webkitpy/test/runner.py:
1358         (_test_name):
1359         (Runner.all_test_names):
1360         (_Worker.handle):
1361
1362 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1363
1364         test-webkitpy: handle failures properly when running in parallel
1365         https://bugs.webkit.org/show_bug.cgi?id=91416
1366
1367         Reviewed by Tony Chang.
1368
1369         It turns out that unittest.TestResults contain a handle to the
1370         test method itself, which isn't picklable; it's sufficient to just
1371         store the test name instead of the actual method. By doing so
1372         we can move the test_name() method from the printer to the
1373         runner where it belongs (so the printer is less dependent on the
1374         unittest framework's data structures).
1375
1376         This change should really have a test but I don't know how to
1377         write one that properly captures the behavior and won't cause
1378         test-webkitpy itself to fail. I've verified the fix by hand, at
1379         least, in the meantime.
1380
1381         * Scripts/webkitpy/test/printer.py:
1382         (Printer.__init__):
1383         (Printer.print_result):
1384         * Scripts/webkitpy/test/runner.py:
1385         (_test_name):
1386         (Runner.all_test_names):
1387         (_Worker.handle):
1388
1389 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1390
1391         test-webkitpy: run tests in parallel
1392         https://bugs.webkit.org/show_bug.cgi?id=91294
1393
1394         Reviewed by Ojan Vafai.
1395
1396         This change adds support for running tests in parallel. This is
1397         not yet on by default, since the logging isn't very pretty w/
1398         parallel tests.
1399
1400         Also, there are some (multiprocessing-related) tests that can't be
1401         run in parallel and so we skip them in that situation; I need to
1402         come up with a mechanism for dealing with this, since you
1403         apparently can't use multiprocessing as both a parent and a
1404         child process.
1405
1406         * Scripts/webkitpy/test/finder.py:
1407         (Finder.find_names):
1408         (Finder._default_names):
1409         * Scripts/webkitpy/test/main.py:
1410         (Tester._parse_args):
1411         (Tester.run):
1412         * Scripts/webkitpy/test/main_unittest.py:
1413         (TesterTest.test_no_tests_found):
1414         * Scripts/webkitpy/test/runner.py:
1415         (Runner.run):
1416         * Scripts/webkitpy/test/runner_unittest.py:
1417         (RunnerTest.test_regular):
1418         (RunnerTest.test_verbose):
1419         (RunnerTest.test_timing):
1420
1421 2012-07-16  Dirk Pranke  <dpranke@chromium.org>
1422
1423         test-webkitpy: use message pools
1424         https://bugs.webkit.org/show_bug.cgi?id=91292
1425
1426         Reviewed by Ojan Vafai.
1427
1428         Restructure the test-running code to be message-driven and
1429         use a MessagePool; note that this does not yet actually run the
1430         tests in parallel.
1431
1432         Also clean up the unit tests so that the fake loader is passed
1433         to the _Worker properly, and reduce a lot of the cut&pasted code
1434         in the tests.
1435
1436         No functional changes; covered by existing tests.
1437
1438         * Scripts/webkitpy/test/printer.py:
1439         (Printer.print_started_test):
1440         (Printer.print_finished_test):
1441         * Scripts/webkitpy/test/runner.py:
1442         (Runner.__init__):
1443         (Runner.run):
1444         (Runner.handle):
1445         (_Worker):
1446         (_Worker.__init__):
1447         (_Worker.handle):
1448         * Scripts/webkitpy/test/runner_unittest.py:
1449         (RunnerTest.assert_run):
1450         (RunnerTest.test_regular):
1451         (RunnerTest.test_verbose):
1452         (RunnerTest.test_timing):
1453
1454 2012-07-16  Simon Hausmann  <simon.hausmann@nokia.com>
1455
1456         [Qt] Fix Qt5 module dependencies
1457         https://bugs.webkit.org/show_bug.cgi?id=91388
1458
1459         Reviewed by Laszlo Gombos.
1460
1461         Removed ancient qtscript and qtxmlpatterns module dependencies.
1462
1463         * qmake/qt_webkit.pri:
1464
1465 2012-07-16  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1466
1467         run-webkit-test returns 254 at exit even when it succeeds
1468         https://bugs.webkit.org/show_bug.cgi?id=91370
1469
1470         Reviewed by Simon Hausmann.
1471
1472         sys.exit() is implemented by raising SystemExit which inherits from
1473         BaseException. Since we are catching BaseException, run_webkit_tests.py
1474         was always returning an error code.
1475
1476         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1477
1478 2012-07-15  Carlos Garcia Campos  <cgarcia@igalia.com>
1479
1480         Unreviewed. Fix make distcheck.
1481
1482         * GNUmakefile.am: Add missing file to compilation.
1483
1484 2012-07-15  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1485
1486         Unreviewed. Add EFLWebKit2PublicAPI, EFLWebKit2PlatformSpecific to watch list.
1487
1488         * Scripts/webkitpy/common/config/watchlist:
1489
1490 2012-07-15  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1491
1492         Unreviewed. Add gyuyoung.kim@samsung.com to watch list.
1493
1494         * Scripts/webkitpy/common/config/watchlist:
1495
1496 2012-07-15  Joseph Pecoraro  <pecoraro@apple.com>
1497
1498         Windowless WebView not firing JavaScript load event if there is a media element
1499         https://bugs.webkit.org/show_bug.cgi?id=91331
1500
1501         Reviewed by Eric Carlson.
1502
1503         Test a windowless WebView loading a page with a media element. The load
1504         event should happen on the next spin of the run loop, but we spin check
1505         the page, with a timeout of 250ms.
1506
1507         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1508         * TestWebKitAPI/Tests/mac/WindowlessWebViewWithMedia.html: Added.
1509         * TestWebKitAPI/Tests/mac/WindowlessWebViewWithMedia.mm: Added.
1510         (-[WindowlessWebViewWithMediaFrameLoadDelegate webView:didFinishLoadForFrame:]):
1511         (TestWebKitAPI::spinLoop): Spin check with timeout. Runs a block each spin to early bail.
1512         (TestWebKitAPI::TEST):
1513
1514 2012-07-14  Benjamin Poulain  <bpoulain@apple.com>
1515
1516         [Mac] Do not try to update the cache model for every WebPreferences change
1517         https://bugs.webkit.org/show_bug.cgi?id=91302
1518
1519         Reviewed by Joseph Pecoraro.
1520
1521         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1522         * TestWebKitAPI/Tests/mac/SetAndUpdateCacheModel.mm: Added.
1523         (TestWebKitAPI):
1524         (TestWebKitAPI::TEST):
1525
1526 2012-07-14  Sheriff Bot  <webkit.review.bot@gmail.com>
1527
1528         Unreviewed, rolling out r122665.
1529         http://trac.webkit.org/changeset/122665
1530         https://bugs.webkit.org/show_bug.cgi?id=91321
1531
1532         Broke Mac builds (Requested by rniwa on #webkit).
1533
1534         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1535         * TestWebKitAPI/Tests/mac/SetAndUpdateCacheModel.mm: Removed.
1536
1537 2012-07-13  Benjamin Poulain  <bpoulain@apple.com>
1538
1539         [Mac] Do not try to update the cache model for every WebPreferences change
1540         https://bugs.webkit.org/show_bug.cgi?id=91302
1541
1542         Reviewed by Joseph Pecoraro.
1543
1544         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1545         * TestWebKitAPI/Tests/mac/SetAndUpdateCacheModel.mm: Added.
1546         (TestWebKitAPI):
1547         (TestWebKitAPI::TEST):
1548
1549 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1550
1551         run-webkit-test outputs the wrong number of tests executed when some are skipped.
1552         https://bugs.webkit.org/show_bug.cgi?id=89894
1553
1554         Reviewed by Ojan Vafai.
1555
1556         Fix the logging of the actual number of tests run so that tests
1557         that are skipped aren't included.
1558
1559         Also revamp the 'expected' output so we distinguish the number
1560         of tests found from the number of tests run (to account for
1561         --repeat-each and --iterations).
1562
1563         Covered by existing tests.
1564
1565         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1566         (Manager.prepare_lists_and_print_output):
1567         (Manager._log_num_workers):
1568         (Manager.run):
1569         (Manager._print_result_summary):
1570         * Scripts/webkitpy/layout_tests/models/result_summary.py:
1571         (ResultSummary.__init__):
1572         (ResultSummary.add):
1573         * Scripts/webkitpy/layout_tests/views/printing.py:
1574         (Printer.print_one_line_summary):
1575         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1576         (Testprinter.test_print_one_line_summary):
1577
1578 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1579
1580         nrwt: actually print the exception name and message for otherwise unhandled exceptions
1581         https://bugs.webkit.org/show_bug.cgi?id=91305
1582
1583         Reviewed by Adam Barth.
1584
1585         Two more places where I was printing the stack trace but not the
1586         exception itself :(. These two spots can't easily be
1587         unit-tested, but I tested them by hand.
1588
1589         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1590         (run):
1591         (main):
1592
1593 2012-07-13  Josh Hawn  <jhawn@apple.com>
1594
1595         Fix for WebContext::getWebCoreStatistics() causes crash if no m_process
1596         https://bugs.webkit.org/show_bug.cgi?id=91116
1597
1598         Reviewed by Simon Fraser.
1599
1600         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1601             Added new test file.
1602         * TestWebKitAPI/Tests/WebKit2/WebCoreStatisticsWithNoWebProcess.cpp: Added.
1603         (TestWebKitAPI::wkContextGetStatisticsCallback):
1604             Tests that callback function receives an error.
1605         (TestWebKitAPI::TEST):
1606             Creates a dummy web context object (no web process).
1607             Calls WKContextGetStatistics with the web context and test callback.
1608             The test callback should get an expected error.
1609
1610 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1611
1612         test-webkitpy: move printing-related code out of the runner
1613         https://bugs.webkit.org/show_bug.cgi?id=91289
1614
1615         Reviewed by Ryosuke Niwa.
1616
1617         More refactoring ... this moves all printing-related stuff out
1618         of runner.py and into printer.py.
1619
1620         No functional changes; covered by existing tests.
1621
1622         * Scripts/webkitpy/test/main.py:
1623         (Tester._run_tests):
1624         * Scripts/webkitpy/test/printer.py:
1625         (Printer.__init__):
1626         (Printer):
1627         (Printer.test_name):
1628         (Printer.print_started_test):
1629         (Printer.print_finished_test):
1630         (Printer.print_result):
1631         * Scripts/webkitpy/test/runner.py:
1632         (Runner.__init__):
1633         (Runner.all_test_names):
1634         (Runner.run):
1635         * Scripts/webkitpy/test/runner_unittest.py:
1636         (RunnerTest.test_regular):
1637         (RunnerTest.test_verbose):
1638         (RunnerTest.test_timing):
1639
1640 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1641
1642         webkitpy: split printing/logging code for test-webkitpy out into a new class
1643         https://bugs.webkit.org/show_bug.cgi?id=91282
1644
1645         Reviewed by Ojan Vafai.
1646
1647         This patch is the first step at splitting all of the
1648         printing/logging code out separately from the actual
1649         test-running code.
1650
1651         This is just moving stuff around; no new functionality and no
1652         new tests needed.
1653
1654         * Scripts/webkitpy/test/finder_unittest.py:
1655         (FinderTest.setUp):
1656         * Scripts/webkitpy/test/main.py:
1657         (Tester.__init__):
1658         (Tester._parse_args):
1659         (Tester.run):
1660         (Tester._run_tests):
1661         (Tester._log_exception):
1662         * Scripts/webkitpy/test/main_unittest.py:
1663         (TesterTest.test_no_tests_found):
1664         * Scripts/webkitpy/test/printer.py: Added.
1665         (Printer):
1666         (Printer.__init__):
1667         (Printer.configure):
1668         (Printer.configure.filter):
1669         (_CaptureAndPassThroughStream):
1670         (_CaptureAndPassThroughStream.__init__):
1671         (_CaptureAndPassThroughStream.write):
1672         (_CaptureAndPassThroughStream._message_is_from_pdb):
1673         (_CaptureAndPassThroughStream.flush):
1674         (_CaptureAndPassThroughStream.getvalue):
1675
1676 2012-07-13  James Simonsen  <simonjam@chromium.org>
1677
1678         [Navigation Timing] Imported W3C tests contain duplicates and are DOS formatted
1679         https://bugs.webkit.org/show_bug.cgi?id=91184
1680
1681         Reviewed by Adam Barth.
1682
1683         The upstream 'html5' tests are just duplicates of the 'html' tests.
1684
1685         * Scripts/import-w3c-performance-wg-tests:
1686
1687 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1688
1689         webkitpy: hide yield_to_caller from callers in MessagePool :)
1690         https://bugs.webkit.org/show_bug.cgi?id=91269
1691
1692         Reviewed by Adam Barth.
1693
1694         yield_to_caller() was an optimization/hack to allow us to run
1695         both manager and worker in a single process/loop without
1696         starving the manager while the worker is running tests. The
1697         worker was required to call yield_to_caller() periodically. It
1698         turns out that I can get equivalent responsiveness by yielding
1699         inside the MessagePool every time the worker posts a message, and this
1700         allows me to no longer need the worker to call the routine. Thus
1701         I rename yield_to_caller() to _yield_to_manager() to be a little
1702         clearer about its purpose.
1703
1704         Tested by existing tests.
1705
1706         * Scripts/webkitpy/common/message_pool.py:
1707         (_Worker.run):
1708         (_Worker.post):
1709         (_Worker._yield_to_manager):
1710         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1711         (Worker.handle):
1712
1713 2012-07-13  Adam Barth  <abarth@webkit.org>
1714
1715         EWSTools should be able to build a commit-queue instance from scratch
1716         https://bugs.webkit.org/show_bug.cgi?id=91264
1717
1718         Reviewed by Eric Seidel.
1719
1720         I've been using this script to build commit-queue instances on Google
1721         Compute Engine and it seems to work.
1722
1723         * EWSTools/GoogleComputeEngine/build-commit-queue.sh: Added.
1724
1725 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1726
1727         webkitpy: make worker.start() and worker.stop() optional in the messagepool
1728         https://bugs.webkit.org/show_bug.cgi?id=91170
1729
1730         Reviewed by Ojan Vafai.
1731
1732         test-webkitpy will use messagepool workers that don't actually
1733         have any per-worker state, so they don't need start() and stop()
1734         methods. Now we will only call the methods if they exist; this
1735         means that workers only need to expose a handle() method.
1736
1737         * Scripts/webkitpy/common/message_pool.py:
1738         (_Worker.terminate):
1739         (_Worker.run):
1740
1741 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1742
1743         NRWT doesn't print exceptions
1744         https://bugs.webkit.org/show_bug.cgi?id=91129
1745
1746         Reviewed by Ojan Vafai.
1747
1748         Although we printed exceptions in most cases, if an unexpected
1749         exception (like a runtime error) was raised when creating a
1750         port, we wouldn't. This patch fixes that, and also cleans up
1751         how we were logging exceptions from the workers to be less
1752         verbose.
1753
1754         Because of the corner cases where these errors are occurring,
1755         it's difficult to write automated unit tests for them. I've
1756         tested it quite a bit by hand, though.
1757
1758         * Scripts/webkitpy/common/message_pool.py:
1759         (_MessagePool._close):
1760         (_MessagePool._handle_worker_exception):
1761         (_Worker.run):
1762         (_Worker._raise):
1763         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1764         (Manager._run_tests):
1765         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1766         (main):
1767
1768 2012-07-13  Wei James  <james.wei@intel.com>
1769
1770         enable TestWebKitAPI/webkit_unit_tests apk on x86 android platform by adding abi support
1771         https://bugs.webkit.org/show_bug.cgi?id=91194
1772
1773         Reviewed by Adam Barth.
1774
1775         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1776
1777 2012-07-13  Simon Pena  <spena@igalia.com>
1778
1779         [GTK] Gardening: update API tests skipped list
1780         https://bugs.webkit.org/show_bug.cgi?id=91224
1781
1782         Unreviewed gardening.
1783
1784         Skip "next" and "previous" tests of FindController until bug #91083
1785         is fixed.
1786
1787         * gtk/run-api-tests:
1788         (TestRunner):
1789
1790 2012-07-13  Zeno Albisser  <zeno@webkit.org>
1791
1792         [Qt][WK2] Implement GraphicsSurface for Linux/GLX.
1793         https://bugs.webkit.org/show_bug.cgi?id=90881
1794
1795         Enable GraphicsSurface for Linux based platforms
1796         whenever the Xcomposite extension is available.
1797
1798         Reviewed by Noam Rosenthal.
1799
1800         * qmake/config.tests/libXcomposite/libXcomposite.cpp: Added.
1801         (main):
1802         * qmake/config.tests/libXcomposite/libXcomposite.pro: Added.
1803           Add a configure test to detect Xcomposite extension and
1804           activate GraphicsSurface on linux in case the extension is available.
1805         * qmake/configure.pri:
1806         * qmake/mkspecs/features/features.prf:
1807
1808 2012-07-13  David Grogan  <dgrogan@chromium.org>
1809
1810         nrwt: don't choke when printing invalid utf-8 to stderr
1811         https://bugs.webkit.org/show_bug.cgi?id=91181
1812
1813         Reviewed by Dirk Pranke.
1814
1815         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1816         (TestResultWriter.write_stderr):
1817
1818 2012-07-13  Dirk Pranke  <dpranke@chromium.org>
1819
1820         test-webkitpy: more class renaming cleanup
1821         https://bugs.webkit.org/show_bug.cgi?id=91182
1822
1823         Reviewed by Adam Barth.
1824
1825         More removing of the unnecessary "Test" prefix.
1826
1827         * Scripts/webkitpy/test/finder.py:
1828         (_DirectoryTree):
1829         (Finder.add_tree):
1830         * Scripts/webkitpy/test/main.py:
1831         (Tester._run_tests):
1832         * Scripts/webkitpy/test/runner.py:
1833         (Runner):
1834         * Scripts/webkitpy/test/runner_unittest.py:
1835         (RunnerTest.test_regular):
1836         (RunnerTest.test_verbose):
1837         (RunnerTest.test_timing):
1838
1839 2012-07-12  Christophe Dumez  <christophe.dumez@intel.com>
1840
1841         [WK2][EFL] Facilitate debugging of the Web Process
1842         https://bugs.webkit.org/show_bug.cgi?id=90768
1843
1844         Reviewed by Kenneth Rohde Christiansen.
1845
1846         Add a new --webprocess-cmd-prefix argument to
1847         run-webkit-tests script for EFL port. If provided,
1848         the prefix will be prepended to the command used
1849         to spawn the Web process. This can be used for
1850         debugging purposes with prefixes such as:
1851         "xterm -title renderer -e gdb --args".
1852
1853         * Scripts/webkitpy/layout_tests/port/efl.py:
1854         (EflPort.__init__):
1855         (EflPort.setup_environ_for_server):
1856         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1857         (parse_args):
1858         * WebKitTestRunner/efl/TestControllerEfl.cpp:
1859         (WTR::TestController::platformRunUntil): Implement support for
1860         m_noTimeout timeout value.
1861
1862 2012-07-12  Adam Barth  <abarth@webkit.org>
1863
1864         Fix crash in the commit-queue.  We need to initialize self.port during __init__.
1865
1866         * Scripts/webkitpy/tool/commands/queues.py:
1867         (CommitQueue.__init__):
1868         (CommitQueue.begin_work_queue):
1869
1870 2012-07-12  Dirk Pranke  <dpranke@chromium.org>
1871
1872         test-webkitpy: rename test_finder to finder
1873         https://bugs.webkit.org/show_bug.cgi?id=91175
1874
1875         Reviewed by Adam Barth.
1876
1877         Rename test_finder -> finder, TestFinder -> Finder to remove
1878         some of the stutter in the names.
1879
1880         * Scripts/webkitpy/test/finder.py: Renamed from Tools/Scripts/webkitpy/test/test_finder.py.
1881         * Scripts/webkitpy/test/finder_unittest.py: Renamed from Tools/Scripts/webkitpy/test/test_finder_unittest.py.
1882         * Scripts/webkitpy/test/main.py:
1883         (Tester.__init__):
1884
1885 2012-07-12  Adam Barth  <abarth@webkit.org>
1886
1887         CommitQueue is confused about what port it is using
1888         https://bugs.webkit.org/show_bug.cgi?id=91040
1889
1890         Reviewed by Dirk Pranke.
1891
1892         On EC2, we explicitly pass --port to the commit-queue, but that
1893         requires editing the start-queue.sh script locally on each bot. In
1894         moving to Google Compute Engine, we're try to avoid any local edits to
1895         the EWSTools.
1896
1897         Rather than passing --port to the commit-queue, this patch teaches the
1898         CommitQueue which port its running, which is the approach we use for
1899         the EWS bots.
1900
1901         Mutating tool._deprecated_port is a bit ugly, but it's what we're doing
1902         currently for the EWS bots.
1903
1904         * Scripts/webkitpy/tool/commands/queues.py:
1905         (CommitQueue):
1906         (CommitQueue.begin_work_queue):
1907         (CommitQueue.run_command):
1908         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1909         (CommitQueueTest.test_commit_queue):
1910         (mock_run_webkit_patch):
1911         (test_rollout):
1912         (test_rollout_lands):
1913         (test_manual_reject_during_processing):
1914
1915 2012-07-12  James Simonsen  <simonjam@chromium.org>
1916
1917         [Navigation Timing] Import the W3C Navigation Timing test suite
1918         https://bugs.webkit.org/show_bug.cgi?id=84887
1919
1920         Reviewed by Tony Gentilcore.
1921
1922         * Scripts/import-w3c-performance-wg-tests: Added.
1923
1924 2012-07-12  Adam Barth  <abarth@webkit.org>
1925
1926         Unreviewed.  Nit: git config files use tabs, not spaces.
1927
1928         * EWSTools/cold-boot.sh:
1929
1930 2012-07-12  Kwang Yul Seo  <skyul@company100.net>
1931
1932         Unreviewed. Add Dongsung Huang to the list of contributors. He
1933         has submitted over 30 patches on texture mapper, canvas and image decoders.
1934
1935         * Scripts/webkitpy/common/config/committers.py:
1936
1937 2012-07-12  Kwang Yul Seo  <skyul@company100.net>
1938
1939         Unreviewed. Change my irc nickname to kseo.
1940
1941         * Scripts/webkitpy/common/config/committers.py:
1942
1943 2012-07-12  Dirk Pranke  <dpranke@chromium.org>
1944
1945         webkitpy: clean up logging handlers, lint common.message_pool
1946         https://bugs.webkit.org/show_bug.cgi?id=91152
1947
1948         Reviewed by Ojan Vafai.
1949
1950         The unix implementation of multiprocessing clones any logging
1951         handlers from the parent process into the child; we currently
1952         don't want this behavior in our code, so I was hand-removing the
1953         installed handlers in the child process I knew about. After thinking
1954         about it further, I think it was simpler and safe enough to just
1955         remove all handlers in the child, since the message pool
1956         propagates any message from the child back into the parent.
1957         
1958         We can always change this in the future if it turns out to be an issue.
1959
1960         I'm also fixing a couple of other lint warnings while I'm at it.
1961
1962         * Scripts/webkitpy/common/message_pool.py:
1963         (_MessagePool.__exit__):
1964         (_MessagePool._handle_worker_exception):
1965         (_Worker._set_up_logging):
1966
1967 2012-07-12  Dirk Pranke  <dpranke@chromium.org>
1968
1969         webkitpy: rename manager_worker_broker to message_pool
1970         https://bugs.webkit.org/show_bug.cgi?id=91145
1971
1972         Reviewed by Ojan Vafai.
1973
1974         Since the MessagePool interface is more generic (and simpler)
1975         now and will be reused by test-webkitpy, I'm renaming it and
1976         moving it to webkitpy.common.
1977
1978         * Scripts/webkitpy/common/message_pool.py: Renamed from Tools/Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py.
1979         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1980         (TestRunInterruptedException.__reduce__):
1981         (Manager._run_tests.worker_factory):
1982         (Manager._run_tests):
1983         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1984
1985 2012-07-12  Dirk Pranke  <dpranke@chromium.org>
1986
1987         nrwt crashes saving the output for a platform-specific expected test reference
1988         https://bugs.webkit.org/show_bug.cgi?id=90872
1989
1990         Reviewed by Ojan Vafai.
1991
1992         The expected output for a test is copied alongside the test
1993         itself in the layout-test-results directory; in other words, for
1994         foo/bar-expected.txt sits alongside foo/bar.html even if we're
1995         actually using platform/mac/foo/bar-expected.txt.
1996
1997         Unless the test is a reftest, in which case we would copy the
1998         output to platform/mac/foo/bar-expected.html and set a
1999         'ref_file' parameter in results.json to indicate the path. This
2000         can be useful in the cases where we have multiple references for
2001         a single test or when multiple tests share the same reference.
2002
2003         We found a bug where we weren't creating platform/mac/foo under
2004         the results directory, and so this wasn't actually working.
2005         However, treating reftests differently seems like a bad thing,
2006         so we should probably be consistent. This change puts the
2007         -expected.html next to the test, and reworks test_result_writer
2008         so that we create directories uniformly and consistently.
2009
2010         Note that we weren't catching this problem in unit tests because
2011         the MockFileSystem creates a directory automatically if it
2012         doesn't exist; this was done intentionally for convenience, but
2013         is really a bug and should be fixed; see https://bugs.webkit.org/show_bug.cgi?id=91028.
2014
2015         I have not added additional tests here since fixing that bug
2016         should be sufficient.
2017
2018         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2019         (interpret_test_failures):
2020         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2021         (ResultSummaryTest.test_interpret_test_failures):
2022         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2023         (write_test_result):
2024         (TestResultWriter._write_binary_file):
2025         (TestResultWriter):
2026         (TestResultWriter._write_text_file):
2027         (TestResultWriter.write_output_files):
2028         (TestResultWriter.write_stderr):
2029         (TestResultWriter.write_crash_log):
2030         (TestResultWriter.create_text_diff_and_write_result):
2031         (TestResultWriter.write_image_diff_files):
2032         (write_reftest):
2033         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2034         (EndToEndTest.test_reftest_with_two_notrefs):
2035
2036 2012-07-12  Dirk Pranke  <dpranke@chromium.org>
2037
2038         nrwt: reimplement manager_worker_broker in a much simpler form
2039         https://bugs.webkit.org/show_bug.cgi?id=90513
2040
2041         Reviewed by Ojan Vafai.
2042
2043         This is a wholesale replacement of the MessagePool() implementation
2044         and the other classes in manager_worker_broker.py. All of the
2045         BrokerConnection*, Broker*, etc. classes are gone, and there are now
2046         just a MessagePool class and a _Worker class. Happiness ensues.
2047  
2048         I'm removing manager_worker_broker_unittest.py as well; we get
2049         nearly complete coverage from the integration tests, and will
2050         get more coverage when test-webkitpy moves to use this as well,
2051         so having unit tests seems like unnecessary overhead. (running
2052         coverage numbers with test-webkitpy shows that pretty much the only
2053         uncovered lines are lines that are only run in the child processes,
2054         which coverage doesn't handle at the moment).
2055  
2056         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2057         (_MessagePool.__init__):
2058         (_MessagePool.run):
2059         (_MessagePool._start_workers):
2060         (_MessagePool):
2061         (_MessagePool.wait):
2062         (_MessagePool._close):
2063         (_MessagePool._handle_done):
2064         (_MessagePool._can_pickle):
2065         (_MessagePool._loop):
2066         (WorkerException):
2067         (_Message.__init__):
2068         (_Message.__repr__):
2069         (_Worker):
2070         (_Worker.__init__):
2071         (_Worker.terminate):
2072         (_Worker._close):
2073         (_Worker.run):
2074         (_Worker.post):
2075         (_Worker.yield_to_caller):
2076         (_Worker._post):
2077         (_Worker._raise):
2078         (_Worker._set_up_logging):
2079         (_WorkerLogHandler.__init__):
2080         (_WorkerLogHandler.emit):
2081         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Removed.
2082
2083 2012-07-12  Tony Chang  <tony@chromium.org>
2084
2085         [chromium] Remove drag and drop API methods that are no longer used
2086         https://bugs.webkit.org/show_bug.cgi?id=90996
2087
2088         Reviewed by Adam Barth.
2089
2090         Migrate DRT to use the methods that take modifier keys.
2091
2092         * DumpRenderTree/chromium/EventSender.cpp:
2093         (EventSender::doDragDrop):
2094         (EventSender::doMouseUp):
2095         (EventSender::doMouseMove):
2096         (EventSender::beginDragWithFiles):
2097
2098 2012-07-12  Ojan Vafai  <ojan@chromium.org>
2099
2100         Allow putting ranges in user.py list prompts
2101         https://bugs.webkit.org/show_bug.cgi?id=91115
2102
2103         Reviewed by Adam Barth.
2104
2105         Ranges are inclusive and denoted by a dash. This is useful for rebaselining a whole port
2106         since the items are listed with each port's builders being contiguous.
2107
2108         * Scripts/webkitpy/common/system/user.py:
2109         (User._wait_on_list_response):
2110         * Scripts/webkitpy/common/system/user_unittest.py:
2111         (UserTest.test_prompt_with_multiple_lists.run_prompt_test):
2112         (UserTest.test_prompt_with_multiple_lists):
2113
2114 2012-07-12  Arnaud Renevier  <arno@renevier.net>
2115
2116         [GTK] Implement disableImageLoading in DRT
2117         https://bugs.webkit.org/show_bug.cgi?id=87973
2118
2119         Reviewed by Martin Robinson.
2120
2121         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2122         (resetDefaultsToConsistentValues):
2123         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2124         (LayoutTestController::disableImageLoading):
2125
2126 2012-07-12  Simon Hausmann  <simon.hausmann@nokia.com>
2127
2128         [Qt] Internal symbols are exported on Linux
2129         https://bugs.webkit.org/show_bug.cgi?id=90981
2130
2131         Reviewed by Jocelyn Turcotte.
2132
2133         I should've re-enabled ELF symbol visibility when removing the symbol map in
2134         r106650.
2135
2136         * qmake/mkspecs/features/default_post.prf:
2137
2138 2012-07-11  Mark Rowe  <mrowe@apple.com>
2139
2140         <http://webkit.org/b/91024> Build against the latest SDK when targeting older OS X versions.
2141
2142         Reviewed by Dan Bernstein.
2143
2144         The deployment target is already set to the version that we're targeting, and it's that setting
2145         which determines which functionality from the SDK is available to us.
2146
2147         * DumpRenderTree/mac/Configurations/Base.xcconfig:
2148         * TestWebKitAPI/Configurations/Base.xcconfig:
2149         * WebKitTestRunner/Configurations/Base.xcconfig:
2150
2151 2012-07-11  Mark Rowe  <mrowe@apple.com>
2152
2153         Replace definitions of BUILDING_ON / TARGETING macros with macros that will error when used.
2154
2155         Part of <http://webkit.org/b/91015> Remove BUILDING_ON / TARGETING macros in favor of system availability macros.
2156
2157         Reviewed by Anders Carlsson.
2158
2159         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Remove the macros completely from here since
2160         they're completely unused in TestNetscapePlugIn.
2161         * DumpRenderTree/config.h:
2162
2163 2012-07-11  Mark Rowe  <mrowe@apple.com>
2164
2165         <http://webkit.org/b/91015> Remove BUILDING_ON / TARGETING macros in favor of system availability macros
2166
2167         This removal was handled by a script that translates the relevant macros in to the equivalent checks
2168         using the system availability macros.
2169
2170         Reviewed by Filip Pizlo.
2171
2172         * DumpRenderTree/cf/WebArchiveDumpSupport.cpp:
2173         * DumpRenderTree/mac/CheckedMalloc.cpp:
2174         * DumpRenderTree/mac/DumpRenderTree.mm:
2175         * DumpRenderTree/mac/DumpRenderTreeDraggingInfo.mm:
2176         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2177         * DumpRenderTree/mac/TextInputController.m:
2178         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
2179         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
2180         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2181
2182 2012-07-11  Robert Kroeger  <rjkroege@chromium.org>
2183
2184         Suppress horizontal conversion of PlatformWheelEvents when hasPreciseScrollingDeltas is true
2185         https://bugs.webkit.org/show_bug.cgi?id=89580
2186
2187         WebKit GTK and Chromium Linux force vertical wheel events to
2188         scroll horizontally when over horizontal scroll bars.  This is
2189         undesirable for touchpad scrolling with
2190         hasPreciseScrollingDeltas() == true. Modified DumpRenderTree to
2191         let a layout test specify this attribute so that the change's impact
2192         on scrolling can be tested in a layout test.
2193
2194         Reviewed by Adam Barth.
2195
2196         * DumpRenderTree/chromium/EventSender.cpp:
2197         (EventSender::handleMouseWheel):
2198
2199 2012-07-11  Simon Fraser  <simon.fraser@apple.com>
2200
2201         Fix the build by declaring -isPaginated before use.
2202
2203         * MiniBrowser/mac/BrowserWindowController.m:
2204
2205 2012-07-11  Sheriff Bot  <webkit.review.bot@gmail.com>
2206
2207         Unreviewed, rolling out r122358.
2208         http://trac.webkit.org/changeset/122358
2209         https://bugs.webkit.org/show_bug.cgi?id=91037
2210
2211         Build break on WebKit Win (Requested by hayato on #webkit).
2212
2213         * DumpRenderTree/chromium/EventSender.cpp:
2214         (EventSender::doDragDrop):
2215         (EventSender::doMouseUp):
2216         (EventSender::doMouseMove):
2217         (EventSender::beginDragWithFiles):
2218
2219 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
2220
2221         nrwt: add a MessagePool abstraction that the manager will call to replace the broker
2222         https://bugs.webkit.org/show_bug.cgi?id=90511
2223
2224         Reviewed by Ojan Vafai.
2225
2226         This change introduces the new MessagePool abstraction that will
2227         replace the classes in manager_worker_broker. It is a minimal
2228         interface that tries to follow the conventions in
2229         multiprocessing.Pool and concurrency.futures ... it provides a
2230         context manager and a run() method that sends N messages to M
2231         workers processes (starting workers as necessary) and waits for
2232         them all to complete, handling cleanup as necessary. The caller
2233         is responsible for providing a handle() method to handle
2234         messages received from the workers.
2235         
2236         This interface basically hides all of the multiprocessing logic from
2237         the manager class.
2238
2239         The initial implementation of MessagePool is a simple shim
2240         around the existing broker classes; a subsequent change will
2241         replace all the other classes with a much simpler
2242         implementation.
2243
2244         No additional tests are provided for now; existing tests should
2245         provide adequate coverage, and I will add new unit tests for the
2246         MessagePool class when I replace the existing implementation.
2247
2248         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2249         (TestRunInterruptedException.__reduce__):
2250         (Manager.__init__):
2251         (Manager._run_tests):
2252         (Manager._run_tests.instead):
2253         (Manager.handle):
2254         (Manager._handle_started_test):
2255         (Manager._handle_finished_test_list):
2256         (Manager._handle_finished_test):
2257         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2258         (get):
2259         (_MessagePool):
2260         (_MessagePool.__init__):
2261         (_MessagePool.__enter__):
2262         (_MessagePool.__exit__):
2263         (_MessagePool.run):
2264         (_MessagePool.wait):
2265         (_MessagePool.is_done):
2266         (_MessagePool._worker_is_done):
2267         (_MessagePool._close):
2268         (_MessagePool.handle_done):
2269         (_MessagePool.handle_started_test):
2270         (_MessagePool.handle_finished_test):
2271         (_MessagePool.handle_finished_test_list):
2272         (_MessagePool.handle_exception):
2273         (_MessagePool._log_messages):
2274         (_MessagePool._handle_worker_exception):
2275         (_WorkerState):
2276         (_WorkerState.for):
2277         (_WorkerState.__init__):
2278         (_WorkerState.__repr__):
2279         (_get_broker):
2280         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2281         (make_broker):
2282
2283 2012-07-11  Simon Fraser  <simon.fraser@apple.com>
2284
2285         Add an option to enter paginated mode in MiniBrowser
2286         https://bugs.webkit.org/show_bug.cgi?id=91035
2287
2288         Reviewed by Dan Bernstein.
2289
2290         Add an item to the Debug menu for MiniBrowser that puts
2291         the web view into paginated mode. For now, we only
2292         do left-to-right pagination, with a fixed column width.
2293
2294         * MiniBrowser/mac/BrowserWindowController.h:
2295         * MiniBrowser/mac/BrowserWindowController.m:
2296         (-[BrowserWindowController validateMenuItem:]): Update the checked
2297         state of the menu item.
2298         (-[BrowserWindowController isPaginated]): Return YES if paginated.
2299         (-[BrowserWindowController togglePaginationMode:]): Toggle in
2300         and out of pagination mode.
2301         * MiniBrowser/mac/MainMenu.xib: Add the debug menu item.
2302
2303 2012-07-11  Adam Barth  <abarth@webkit.org>
2304
2305         commit-queue instances on Compute Engine are missing git-svn
2306         https://bugs.webkit.org/show_bug.cgi?id=91034
2307
2308         Reviewed by Eric Seidel.
2309
2310         git-svn is needed to actually commit to SVN from a git working copy.
2311
2312         * EWSTools/cold-boot.sh:
2313
2314 2012-07-11  Dirk Pranke  <dpranke@chromium.org>
2315
2316         webkitpy: lint code in webkitpy.layout_tests.models
2317         https://bugs.webkit.org/show_bug.cgi?id=90416
2318
2319         Reviewed by Ojan Vafai.
2320
2321         Cleaning up errors reported from lint-webkitpy.
2322
2323         Also, suppress the warnings about wildcard imports in pylintrc;
2324         we have nothing particularly against them.
2325
2326         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
2327         (TestConfigurationConverter.combinations):
2328         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
2329         (TestConfigurationTest.test_hash.query_unknown_key):
2330         (TestConfigurationTest.test_eq):
2331         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2332         (ParseError.__init__):
2333         (TestExpectationLine.__init__):
2334         (TestExpectationsModel.get_expectations_string):
2335         (TestExpectationsModel):
2336         (TestExpectationsModel.expectation_to_string):
2337         (TestExpectationsModel.add_expectation_line):
2338         (TestExpectationsModel._clear_expectations_for_test):
2339         (TestExpectationsModel._remove_from_sets):
2340         (TestExpectations.get_expectations_string):
2341         (TestExpectations.expectation_to_string):
2342         (TestExpectations._report_warnings):
2343         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2344         (Base.__init__):
2345         (parse_exp):
2346         (SkippedTests.check):
2347         (TestExpectationParserTests.test_parse_empty_string):
2348         * Scripts/webkitpy/layout_tests/models/test_failures.py:
2349         (FailureTimeout.__init__):
2350         (FailureCrash.__init__):
2351         (FailureImageHashMismatch.__init__):
2352         (FailureReftestMismatch.__init__):
2353         (FailureReftestMismatchDidNotOccur.__init__):
2354         (FailureReftestNoImagesGenerated.__init__):
2355         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py:
2356         (TestFailuresTest.test_unknown_failure_type.UnknownFailure.message):
2357         (TestFailuresTest.test_unknown_failure_type):
2358         (TestFailuresTest):
2359         (TestFailuresTest.test_message_is_virtual):
2360         * Scripts/webkitpy/layout_tests/models/test_results.py:
2361         (TestResult.loads):
2362         (TestResult.has_failure_matching_types):
2363         * Scripts/webkitpy/pylintrc:
2364
2365 2012-07-11  Dirk Pranke  <dpranke@chromium.org>
2366
2367         nrwt: clean up names in worker.py
2368         https://bugs.webkit.org/show_bug.cgi?id=90510
2369
2370         Reviewed by Ojan Vafai.
2371
2372         This is the last patch in the series of refactoring worker.py;
2373         all this does is change some names of methods, instance
2374         variables, and method parameters to be clearer (it also changes
2375         some code in manager.py and manager_worker.py to be consistent).
2376
2377         There are no functional changes in this patch and the existing
2378         tests should cover everything.
2379
2380         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2381         (Manager.__init__):
2382         (Manager.prepare_lists_and_print_output):
2383         (Manager._run_tests.worker_factory):
2384         (Manager._run_tests):
2385         (Manager._show_results_html_file):
2386         (Manager.handle_finished_test_list):
2387         (_WorkerState.__init__):
2388         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2389         (_Broker.post_message):
2390         (_Broker._dispatch_message):
2391         (AbstractWorker.__init__):
2392         (AbstractWorker.run):
2393         (AbstractWorker.yield_to_caller):
2394         (AbstractWorker.post):
2395         (_WorkerConnection.__init__):
2396         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2397         (_TestWorker.__init__):
2398         (_TestWorker.start):
2399         (_TestWorker.handle):
2400         (_TestWorker.stop):
2401         (_TestsMixin):
2402         (_TestsMixin.test_name):
2403         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2404         (Worker.__init__):
2405         (Worker.__del__):
2406         (Worker.start):
2407         (Worker.handle):
2408         (Worker._run_test):
2409         (Worker.stop):
2410         (Worker._timeout):
2411         (Worker._kill_driver):
2412         (Worker._run_test_with_timeout):
2413         (Worker._clean_up_after_test):
2414         (Worker._run_test_in_another_thread.SingleTestThread.run):
2415         (Worker._run_test_in_this_thread):
2416         (Worker._run_single_test):
2417
2418 2012-07-11  Adam Barth  <abarth@webkit.org>
2419
2420         The commit-queue needs some extra git config to be able to commit
2421         https://bugs.webkit.org/show_bug.cgi?id=91025
2422
2423         Reviewed by Eric Seidel.
2424
2425         In order for the commit-queue to actually commit, it needs to know the
2426         location of the SVN server.
2427
2428         * EWSTools/cold-boot.sh:
2429
2430 2012-07-11  Adam Barth  <abarth@webkit.org>
2431
2432         Teach EWSTools how to configure SVN auth credentials
2433         https://bugs.webkit.org/show_bug.cgi?id=91021
2434
2435         Reviewed by Eric Seidel.
2436
2437         To move the commit-queue over to Google Compute Engine, we need a way
2438         to configure the commit-queue's credentials during the machine build
2439         process. There doesn't seem to be a nice command line way of
2440         configuring SVN auth credentials. I tried doing a bogus commit and
2441         supplying the credentials that way, but that's super hacky.
2442
2443         The approach in this patch is to write the config file directly. The
2444         format of these configurations files is documented in the SVN book, so
2445         this approach doesn't see too sketchy.
2446
2447         * EWSTools/configure-svn-auth.sh: Added.
2448
2449 2012-07-11  Ojan Vafai  <ojan@chromium.org>
2450
2451         Dedupe suffixes passed to webkit-patch rebaseline
2452         https://bugs.webkit.org/show_bug.cgi?id=91017
2453
2454         Reviewed by Dirk Pranke.
2455
2456         * Scripts/webkitpy/tool/commands/rebaseline.py:
2457         (Rebaseline._suffixes_to_update):
2458         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2459         (test_rebaseline_multiple_builders_and_tests_command_line):
2460
2461 2012-07-11  Kevin Ollivier  <kevino@theolliviers.com>
2462
2463         [wx] Unreviewed build fix. Add new directories and a new LayoutTestController method.
2464
2465         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2466         (LayoutTestController::setStorageDatabaseIdleInterval):
2467         * waf/build/settings.py:
2468
2469 2012-07-11  Tony Chang  <tony@chromium.org>
2470
2471         [chromium] Remove drag and drop API methods that are no longer used
2472         https://bugs.webkit.org/show_bug.cgi?id=90996
2473
2474         Reviewed by Adam Barth.
2475
2476         Migrate DRT to use the methods that take modifier keys.
2477
2478         * DumpRenderTree/chromium/EventSender.cpp:
2479         (EventSender::doDragDrop):
2480         (EventSender::doMouseUp):
2481         (EventSender::doMouseMove):
2482         (EventSender::beginDragWithFiles):
2483
2484 2012-07-11  Mark Rowe  <mrowe@apple.com>
2485
2486         Add a Mountain Lion version of libWebKitSystemInterface.a.
2487
2488         Reviewed by John Sullivan.
2489
2490         * Scripts/copy-webkitlibraries-to-product-directory: Include libWebKitSystemInterfaceMountainLion.a in the list of libraries to copy.
2491
2492 2012-07-09  Mark Rowe  <mrowe@apple.com>
2493
2494         <http://webkit.org/b/90835> Teach bisect-builds to work with a Safari.app that has entitlements.
2495
2496         Reviewed by Dan Bernstein.
2497
2498         * Scripts/bisect-builds: Use safariPathFromSafariBundle to determine which binary within the application
2499         should be invoked.
2500
2501 2012-07-11  Zan Dobersek  <zandobersek@gmail.com>
2502
2503         [Gtk] fast/events/keydown-function-keys.html is failing
2504         https://bugs.webkit.org/show_bug.cgi?id=90891
2505
2506         Reviewed by Martin Robinson.
2507
2508         Work around the context menu being shown on F10 key being pressed by
2509         unbiding the key when running tests in DumpRenderTree. The problem
2510         appears when using a recent version of the xkeyboard-config package.
2511
2512         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2513         (setDefaultsToConsistentStateValuesForTesting):
2514
2515 2012-07-11  No'am Rosenthal  <noam.rosenthal@nokia.com>
2516
2517         [Qt] QRawWebView should notify when rendering is done, so that pixel results can be grabbed at the appropriate moment.
2518         https://bugs.webkit.org/show_bug.cgi?id=90641
2519
2520         Reviewed by Jocelyn Turcotte.
2521
2522         * MiniBrowser/qt/raw/View.h: Comment used old class name (WKView).
2523         (View):
2524
2525 2012-07-11  Min Qin  <qinmin@chromium.org>
2526
2527         [Android] sending an extra to the DRT apk so that DRT can be run in a seperate thread
2528         https://bugs.webkit.org/show_bug.cgi?id=90831
2529
2530         Reviewed by Adam Barth.
2531
2532         On android, DRT needs to run in a background thread to avoid ANR.
2533         However, the java tests are running on UI thread by default.
2534         We need to send an intent extra to the apk so that it can run on a sub thread.
2535
2536         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2537         (ChromiumAndroidDriver._start_once):
2538
2539 2012-07-11  Csaba Osztrogonác  <ossy@webkit.org>
2540
2541         [Qt] REGRESSION(r107171): Fix --timeout option of Qt's DRT
2542         https://bugs.webkit.org/show_bug.cgi?id=90966
2543
2544         Reviewed by Ryosuke Niwa.
2545
2546         * DumpRenderTree/qt/main.cpp:
2547         (main): Don't remove the argument, because takeOptionValue() did it before.
2548
2549 2012-07-11  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
2550
2551         First commit; moving myself to commiters.
2552
2553         Unreviewed.
2554
2555         * Scripts/webkitpy/common/config/committers.py:
2556
2557 2012-07-11  Balazs Ankes  <bank@inf.u-szeged.hu>
2558
2559         webkit-patch land should automatically add svn:mime-type for .png files
2560         https://bugs.webkit.org/show_bug.cgi?id=75825
2561
2562         Reviewed by Dirk Pranke.
2563
2564         Refactor the png.py to avoid code duplication.
2565
2566         * Scripts/webkitpy/common/checksvnconfigfile.py: Added.
2567         (check):
2568         (config_file_path):
2569         (errorstr_autoprop):
2570         (errorstr_png):
2571         * Scripts/webkitpy/style/checkers/png.py:
2572         (PNGChecker.check):
2573         * Scripts/webkitpy/tool/commands/download.py:
2574         (Land):
2575         * Scripts/webkitpy/tool/steps/__init__.py:
2576         * Scripts/webkitpy/tool/steps/addsvnmimetypeforpng.py: Added.
2577         (AddSvnMimetypeForPng):
2578         (AddSvnMimetypeForPng.__init__):
2579         (AddSvnMimetypeForPng.run):
2580         (AddSvnMimetypeForPng._check_pngs):
2581         * Scripts/webkitpy/tool/steps/addsvnmimetypeforpng_unittest.py: Added.
2582         (MockSCMDetector):
2583         (MockSCMDetector.__init__):
2584         (MockSCMDetector.display_name):
2585         (AddSvnMimetypeForPngTest):
2586         (AddSvnMimetypeForPngTest.test_run):
2587
2588 2012-07-11  Hayato Ito  <hayato@chromium.org>
2589
2590         Unreviewed gardening.
2591         One more fix for r122292.
2592
2593         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2594
2595 2012-07-11  Sheriff Bot  <webkit.review.bot@gmail.com>
2596
2597         Unreviewed, rolling out r122301.
2598         http://trac.webkit.org/changeset/122301
2599         https://bugs.webkit.org/show_bug.cgi?id=90947
2600
2601         Android builder started to fail (Requested by hayato on
2602         #webkit).
2603
2604         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2605
2606 2012-07-10  Adam Barth  <abarth@webkit.org>
2607
2608         [Chromium] Merge final nits to DumpRenderTree.gyp for Android
2609         https://bugs.webkit.org/show_bug.cgi?id=90920
2610
2611         Reviewed by Tony Chang.
2612
2613         This patch contains the last few small changes to DumpRenderTree.gyp
2614         from the chromium-android branch. After this change, this file will be
2615         fully merged.
2616
2617         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2618
2619 2012-07-10  Adam Barth  <abarth@webkit.org>
2620
2621         [Chromium-Android] Add apk test targets for webkit_unit_tests and TestWebKitAPI
2622         https://bugs.webkit.org/show_bug.cgi?id=90918
2623
2624         Reviewed by Tony Chang.
2625
2626         The rules are similar to what we have added for DumpRenderTree apk.
2627         All references to gtest_target_type can be removed once we enable APK
2628         tests on the all bots.
2629
2630         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2631
2632 2012-07-10  Kenneth Russell  <kbr@google.com>
2633
2634         Add --no-build option to perf test runner
2635         https://bugs.webkit.org/show_bug.cgi?id=90916
2636
2637         Reviewed by Ryosuke Niwa.
2638
2639         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2640         (PerfTestsRunner._parse_args):
2641             Support --no-build option to perf tests, as in layout tests.
2642
2643 2012-07-10  Adam Barth  <abarth@webkit.org>
2644
2645         EWSTools should be able to build a chromium-ews bot from scratch
2646         https://bugs.webkit.org/show_bug.cgi?id=90912
2647
2648         Reviewed by Eric Seidel.
2649
2650         I've been using this script to kick off the build process for the
2651         chromium-ews bots on Google Compute Engine.
2652
2653         * EWSTools/GoogleComputeEngine: Added.
2654         * EWSTools/GoogleComputeEngine/build-chromium-ews.sh: Added.
2655
2656 2012-07-10  Alec Flett  <alecflett@chromium.org>
2657
2658         Add alecflett to the list of committers
2659         https://bugs.webkit.org/show_bug.cgi?id=90903
2660
2661         Reviewed by Ojan Vafai.
2662
2663         * Scripts/webkitpy/common/config/committers.py:
2664
2665 2012-07-10  Xianzhu Wang  <wangxianzhu@chromium.org>
2666
2667         Unreviewed fix of webkitpy unittest break about chromium_android.py
2668         caused by r122251.
2669
2670         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2671         (ChromiumAndroidDriverTest.test_cmd_line):
2672
2673 2012-07-10  Ojan Vafai  <ojan@chromium.org>
2674
2675         webkit-patch rebaseline doesn't work for audio/pixel tests
2676         https://bugs.webkit.org/show_bug.cgi?id=90905
2677
2678         Reviewed by Adam Barth.
2679
2680         We just need to pass the correct suffix list to rebaseline-test-internal
2681         and optimize-baselines. By default, pass all suffixes.
2682
2683         * Scripts/webkitpy/tool/commands/rebaseline.py:
2684         (Rebaseline):
2685         (Rebaseline.__init__):
2686         (Rebaseline._suffixes_to_update):
2687         (Rebaseline.execute):
2688         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2689         (test_rebaseline_multiple_builders_and_tests_command_line):
2690
2691 2012-07-10  Adam Barth  <abarth@webkit.org>
2692
2693         Remove LayoutTestController and WebKitTestRunner support for Hixie76 WebSockets
2694         https://bugs.webkit.org/show_bug.cgi?id=90853
2695
2696         Reviewed by Eric Seidel.
2697
2698         WebKit no longer implements the Hixie76 version of the WebSocket
2699         protocol. We don't need to support it in our test framework.
2700
2701         * DumpRenderTree/chromium/LayoutTestController.cpp:
2702         (LayoutTestController::overridePreference):
2703         * DumpRenderTree/chromium/WebPreferences.cpp:
2704         (WebPreferences::reset):
2705         (WebPreferences::applyTo):
2706         * DumpRenderTree/chromium/WebPreferences.h:
2707         (WebPreferences):
2708         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2709         (resetDefaultsToConsistentValues):
2710         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2711         (LayoutTestController::overridePreference):
2712         * DumpRenderTree/mac/DumpRenderTree.mm:
2713         (resetDefaultsToConsistentValues):
2714         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2715         (WebCore::WebPage::resetSettings):
2716         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2717         (LayoutTestController::overridePreference):
2718         * DumpRenderTree/win/DumpRenderTree.cpp:
2719         (resetDefaultsToConsistentValues):
2720
2721 2012-07-10  Xianzhu Wang  <wangxianzhu@chromium.org>
2722
2723         [Chromium-Android] Fix typos in chromium_android.py in r151492
2724         https://bugs.webkit.org/show_bug.cgi?id=90904
2725
2726         Reviewed by Ojan Vafai.
2727
2728         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2729         (ChromiumAndroidPort.setup_test_run):
2730
2731 2012-07-10  Xianzhu Wang  <wangxianzhu@chromium.org>
2732
2733         [Chromium-Android] Use setup_test_runner() instead of start_helper() to setup test environment
2734         https://bugs.webkit.org/show_bug.cgi?id=90894
2735
2736         Reviewed by Adam Barth.
2737
2738         start_helper() is actually start_pixel_test_helper() since r115601 (bug 81729).
2739         Should use setup_test_runner() to setup test environment for chromium-android.
2740
2741         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2742         (ChromiumAndroidPort.setup_test_run): Renamed from start_helper(). Added cache cleanup code.
2743         (ChromiumAndroidPort.clean_up_test_run): Renamed from stop_helper().
2744         (ChromiumAndroidPort._path_to_helper): Returns None as we don't have a helper now.
2745         (ChromiumAndroidPort):
2746         (ChromiumAndroidPort._path_to_forwarder): Original _path_to_helper().
2747         (ChromiumAndroidPort._push_executable):
2748         (ChromiumAndroidDriver.__init__):
2749         (ChromiumAndroidDriver.cmd_line):
2750
2751 2012-07-09  Ojan Vafai  <ojan@chromium.org>
2752
2753         Improve webkit-patch rebaseline to work for more cases
2754         https://bugs.webkit.org/show_bug.cgi?id=90504
2755
2756         Reviewed by Adam Barth.
2757
2758         -Makes it work for the build.chromium.org bots.
2759         -Lets you rebaseline all builders instead of just one.
2760         -Lets you pass in the builders or tests to rebaseline.
2761
2762         * Scripts/webkitpy/common/host.py:
2763         (Host.buildbot_for_builder_name):
2764         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2765         (Builder.__init__):
2766         (Builder.latest_layout_test_results):
2767         Provide a way to get to the LayoutTestResults of the latest build.
2768         Most of the time we only need to get the latest one and the Chromium bots
2769         only expose the full_results.json file for the latest build.
2770
2771         (Builder):
2772         (Builder._fetch_file_from_results):
2773         (Builder.fetch_layout_test_results):
2774         Move these functions into Builder so that Builder and Build can both
2775         fetch layout test results.
2776
2777         (Build.results_zip_url):
2778         (Build.layout_test_results):
2779         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2780         (BuilderTest.test_latest_layout_test_results):
2781         (BuildTest.test_layout_test_results):
2782         * Scripts/webkitpy/common/system/user.py:
2783         (User.prompt_with_multiple_lists):
2784         Prompt with multiple sublists, but still have a global numbering.
2785         This lets the build.chromium.org bots be clearly separated from the
2786         build.webkit.org bots, which helps understand the builder names.
2787
2788         (User):
2789         (User._wait_on_list_response):
2790         (User.prompt_with_list):
2791         * Scripts/webkitpy/common/system/user_unittest.py:
2792         (UserTest.test_prompt_with_multiple_lists):
2793         (UserTest.test_prompt_with_multiple_lists.run_prompt_test):
2794         (UserTest.test_prompt_with_multiple_lists.run_prompt_test.mock_raw_input):
2795         * Scripts/webkitpy/layout_tests/port/base.py:
2796         (Port.is_chromium):
2797         Provide a way to tell if a port is a Chromium port that doesn't involve string
2798         manipulation on the port name or builder name.
2799
2800         * Scripts/webkitpy/layout_tests/port/builders.py:
2801         (builder_path_from_name):
2802         (all_builder_names):
2803         Memoizing here is incorrect because the test override _exact_matches,
2804         so these can return different values. In either case, I'm pretty sure these
2805         are not remotely hot codepaths. I manually inspected all call sites and they're
2806         all in high-level calls (e.g. execute calls for webkit-patch commands) and not
2807         called in a loop.
2808
2809         * Scripts/webkitpy/layout_tests/port/chromium.py:
2810         (ChromiumPort.is_chromium):
2811         * Scripts/webkitpy/tool/commands/rebaseline.py:
2812         (RebaselineTest._results_url):
2813         (AbstractParallelRebaselineCommand.__init__):
2814         (Rebaseline.__init__):
2815         (Rebaseline):
2816         (Rebaseline._builders_to_pull_from):
2817         (Rebaseline._tests_to_update):
2818         (Rebaseline.execute):
2819         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2820         (test_rebaseline.mock_builders_to_pull_from):
2821         (test_rebaseline):
2822         (test_rebaseline_command_line_flags):
2823         (test_rebaseline_multiple_builders):
2824         (test_rebaseline_multiple_builders.mock_builders_to_pull_from):
2825         (test_rebaseline_multiple_builders.mock_tests_to_update):
2826         (test_rebaseline_multiple_builders_and_tests_command_line):
2827         * Scripts/webkitpy/tool/mocktool.py:
2828         (MockTool.irc):
2829         (MockTool):
2830         (MockTool.buildbot_for_builder_name):
2831
2832 2012-07-10  Alice Cheng  <alice_cheng@apple.com>
2833
2834         Editing: Reproducible crasher when pasting a 0x0 image into Mail
2835         https://bugs.webkit.org/show_bug.cgi?id=90640
2836         <rdar://problem/11141920>
2837
2838         Reviewed by Brady Eidson.
2839
2840         Test cases for the patch. 0x0 images don't get a resource representation in the WebArchive, so we need a null check.
2841
2842         * TestWebKitAPI/Tests/mac/0.png: Added.
2843         * TestWebKitAPI/Tests/mac/WebViewCanPasteZeroPng.mm: Added.
2844         (TestWebKitAPI):
2845         (TestWebKitAPI::TEST):
2846
2847 2012-07-10  Leandro Gracia Gil  <leandrogracia@chromium.org>
2848
2849         WebSurroundingText layout tests should use the same code path as the rest of the feature.
2850         https://bugs.webkit.org/show_bug.cgi?id=90807
2851
2852         Reviewed by Adam Barth.
2853
2854         Make the textSurroundingNode method take a pair of point coordinates
2855         instead of a node offset.
2856
2857         * DumpRenderTree/chromium/LayoutTestController.cpp:
2858         (LayoutTestController::textSurroundingNode):
2859         * DumpRenderTree/chromium/LayoutTestController.h:
2860         (LayoutTestController):
2861
2862 2012-07-10  Csaba Osztrogonác  <ossy@webkit.org>
2863
2864         [Qt] REGRESSION(r122175): It broke the Windows build
2865         https://bugs.webkit.org/show_bug.cgi?id=90850
2866
2867         Reviewed by Noam Rosenthal.
2868
2869         * qmake/mkspecs/features/features.prf: Disable USE(3D_GRAPHICS) on Windows temporarily until proper fix.
2870
2871 2012-07-10  János Badics  <jbadics@inf.u-szeged.hu>
2872
2873         [Qt][NRWT] Enable cascaded TestExpectations
2874         https://bugs.webkit.org/show_bug.cgi?id=89108
2875
2876         Reviewed by Dirk Pranke.
2877
2878         * Scripts/webkitpy/layout_tests/port/qt.py:
2879         (QtPort.expectations_files):
2880         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2881         (QtPortTest._assert_expectations_files):
2882         (QtPortTest.test_expectations_files):
2883
2884 2012-07-10  Christophe Dumez  <christophe.dumez@intel.com>
2885
2886         [EFL] WebKit DRT and WTR fail to build due to undefined reference to WTF::MD5::*
2887         https://bugs.webkit.org/show_bug.cgi?id=90868
2888
2889         Unreviewed EFL build fix.
2890
2891         Correct CMake configuration to have EFL's DRT and WebKitTestRunner link
2892         WTF library. This is needed to resolve undefined reference to WTF::MD5::*.
2893
2894         * DumpRenderTree/efl/CMakeLists.txt:
2895         * WebKitTestRunner/PlatformEfl.cmake:
2896
2897 2012-07-10  Adam Barth  <abarth@webkit.org>
2898
2899         LayoutTestController.dumpConfigurationForViewport should move to Internals
2900         https://bugs.webkit.org/show_bug.cgi?id=45652
2901
2902         Reviewed by Eric Seidel.
2903
2904         Remove dumpConfigurationForViewport from LayoutTestController. This API
2905         has been replaced by an API on Internals that does the same thing in a
2906         port-agnostic way.
2907
2908         * DumpRenderTree/LayoutTestController.cpp:
2909         (LayoutTestController::staticFunctions):
2910         * DumpRenderTree/LayoutTestController.h:
2911         (LayoutTestController):
2912         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
2913         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2914         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2915         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2916         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2917         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2918         (LayoutTestController):
2919         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2920         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2921
2922 2012-07-09  Matt Falkenhagen  <falken@chromium.org>
2923
2924         Add ENABLE_DIALOG_ELEMENT and skeleton files
2925         https://bugs.webkit.org/show_bug.cgi?id=90521
2926
2927         Reviewed by Kent Tamura.
2928
2929         * Scripts/webkitperl/FeatureList.pm:
2930
2931 2012-07-09  Kent Tamura  <tkent@chromium.org>
2932
2933         Improve the boilerplate generated by prepare-ChangeLog
2934         https://bugs.webkit.org/show_bug.cgi?id=89560
2935
2936         Reviewed by Ryosuke Niwa.
2937
2938         Produce the following boilerplate:
2939
2940         > 2012-06-20  Kent Tamura  <tkent@chromium.org>
2941         >
2942         >         Need a short description (Oops!).
2943         >         Need the bug URL (Oops!).
2944         >
2945         >         Reviewed by NOBODY (Oops!).
2946         >
2947         >         Additional information of the change such as approach, rationale. Please add per-function descriptions below. (Oops!).
2948         >
2949         >         No new tests (Oops!).
2950
2951         * Scripts/prepare-ChangeLog:
2952         (generateNewChangeLogs): Generate the updated template.
2953         * Scripts/webkitpy/common/checkout/changelog.py:
2954         (ChangeLog.update_with_unreviewed_message):
2955         Update the pattern for the beginning of the boilerplate.
2956         (ChangeLog.set_short_description_and_bug_url):
2957         - Use self._changelog_indent
2958         - Update the pattern for the description line
2959         - Skip the bug boilerplate.
2960         We substitute the specified bug URL to the description boilerplate
2961         because we need to add it even if the bug URL boilerplate is missing.
2962         * Scripts/webkitpy/common/checkout/changelog_unittest.py: Update the boilerplate data.
2963         (test_set_short_description_and_bug_url):
2964         - Update the description boilerplate.
2965         - Add a test for a case with both of the description boilerplace and the
2966           bug URL boilerplate.
2967         * Scripts/webkitpy/tool/steps/preparechangelog_unittest.py:
2968         (PrepareChangeLogTest.test_ensure_bug_url): Update the expected boilerplate.
2969
2970 2012-07-09  No'am Rosenthal  <noam.rosenthal@nokia.com>
2971
2972         Shared code that is guarded with ENABLE(WEBGL) should be guarded with USE()
2973         https://bugs.webkit.org/show_bug.cgi?id=90506
2974
2975         Reviewed by Martin Robinson.
2976
2977         Always enable the WTF_USE_3D_GRAPHICS flag in features.prf if we're compiling with Qt5
2978         or if the OpenGL module is enabled.
2979
2980         * qmake/mkspecs/features/features.prf:
2981
2982 2012-07-09  Yongjun Zhang  <yongjun_zhang@apple.com>
2983
2984         Consider closing unused localStorage database after a timeout.
2985         https://bugs.webkit.org/show_bug.cgi?id=90713
2986
2987         For a localStorage, if there is no active document referencing to it for certain amount of time (300 seconds),
2988         we can close the underlying sqlite database.
2989
2990         Reviewed by Brady Eidson.
2991
2992         * DumpRenderTree/LayoutTestController.cpp: add a JS method setStorageDatabaseIdleInterval to testRunner
2993             to set a different timeout value for closing localStorage database.
2994         (setStorageDatabaseIdleIntervalCallback):
2995         (LayoutTestController::staticFunctions):
2996         * DumpRenderTree/LayoutTestController.h:
2997         (LayoutTestController):
2998         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2999         (LayoutTestController::setStorageDatabaseIdleInterval): add empty method.
3000         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3001         (LayoutTestController::setStorageDatabaseIdleInterval): ditto.
3002         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
3003         (LayoutTestController::setStorageDatabaseIdleInterval): ditto.
3004         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
3005         (LayoutTestController::setStorageDatabaseIdleInterval): ditto.
3006
3007 2012-07-09  Adam Klein  <adamk@chromium.org>
3008
3009         Rename WebCore::WebKitMutationObserver to WebCore::MutationObserver
3010         https://bugs.webkit.org/show_bug.cgi?id=90810
3011
3012         Reviewed by Ojan Vafai.
3013
3014         * DumpRenderTree/gtk/EventSender.cpp:
3015
3016 2012-07-09  Mihai Parparita  <mihaip@chromium.org>
3017
3018         Handle missing results in TestResultsServer better
3019         https://bugs.webkit.org/show_bug.cgi?id=90816
3020
3021         Reviewed by Ojan Vafai.
3022
3023         If we can't load the results JSON, don't try to wrap it with the JSONP
3024         callback invocation.
3025
3026         * TestResultServer/handlers/testfilehandler.py:
3027         (GetFile.get):
3028
3029 2012-07-09  Sheriff Bot  <webkit.review.bot@gmail.com>
3030
3031         Unreviewed, rolling out r122124.
3032         http://trac.webkit.org/changeset/122124
3033         https://bugs.webkit.org/show_bug.cgi?id=90815
3034
3035         It broke NRWT on Qt (Requested by Ossy_HOME on #webkit).
3036
3037         * Scripts/webkitpy/layout_tests/port/qt.py:
3038         (QtPort._skipped_file_search_paths):
3039         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3040         (QtPortTest._assert_skipped_path):
3041         (QtPortTest.test_skipped_file_search_path):
3042
3043 2012-07-09  Zan Dobersek  <zandobersek@gmail.com>
3044
3045         [Gtk] Implement dumpFrameScrollPosition in DumpRenderTree
3046         https://bugs.webkit.org/show_bug.cgi?id=89356
3047
3048         Reviewed by Martin Robinson.
3049
3050         Implement dumpFrameScrollPosition, acquiring the scroll position
3051         through WebKitDOMDOMWindow of the WebKitDOMDocument loaded in the frame
3052         whose scroll position is being dumped.
3053
3054         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3055         (dumpFrameScrollPosition):
3056
3057 2012-07-09  János Badics  <jbadics@inf.u-szeged.hu>
3058
3059         [Qt][NRWT] Enable cascaded TestExpectations
3060         https://bugs.webkit.org/show_bug.cgi?id=89108
3061
3062         Reviewed by Dirk Pranke.
3063
3064         * Scripts/webkitpy/layout_tests/port/qt.py:
3065         (QtPort.expectations_files):
3066         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3067         (QtPortTest._assert_expectations_files):
3068         (QtPortTest.test_expectations_files):
3069
3070 2012-07-09  Dan Bernstein  <mitz@apple.com>
3071
3072         Fixed clean builds of the All target in the DumpRenderTree project.
3073
3074         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Let Xcode know that the
3075         DumpRenderTree Perl Support target depends on the DumpRenderTree target.
3076
3077 2012-07-09  Zan Dobersek  <zandobersek@gmail.com>
3078
3079         [Gtk] Unskip the CSS Regions layout tests
3080         https://bugs.webkit.org/show_bug.cgi?id=90771
3081
3082         Reviewed by Martin Robinson.
3083
3084         Use the DumpRenderTreeSupportGtk method for enabling the CSS Regions,
3085         calling it when resetting the defaults in DumpRenderTree or when the
3086         settings value is overriden through the testRunner.
3087
3088         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3089         (resetDefaultsToConsistentValues):
3090         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3091         (LayoutTestController::overridePreference):
3092
3093 2012-07-09  Sheriff Bot  <webkit.review.bot@gmail.com>
3094
3095         Unreviewed, rolling out r122107.
3096         http://trac.webkit.org/changeset/122107
3097         https://bugs.webkit.org/show_bug.cgi?id=90794
3098
3099         Build failure on Mac debug bots (Requested by falken_ on
3100         #webkit).
3101
3102         * Scripts/webkitperl/FeatureList.pm:
3103
3104 2012-07-09  Matt Falkenhagen  <falken@chromium.org>
3105
3106         Add ENABLE_DIALOG_ELEMENT and skeleton files
3107         https://bugs.webkit.org/show_bug.cgi?id=90521
3108
3109         Reviewed by Kent Tamura.
3110
3111         * Scripts/webkitperl/FeatureList.pm:
3112
3113 2012-07-08  Kent Tamura  <tkent@chromium.org>
3114
3115         Unreviewed, rolling out r122076.
3116         http://trac.webkit.org/changeset/122076
3117         https://bugs.webkit.org/show_bug.cgi?id=89560
3118
3119         Broke sheriffbot rollout
3120
3121         * Scripts/prepare-ChangeLog:
3122         (generateNewChangeLogs):
3123
3124 2012-07-08  Leandro Gracia Gil  <leandrogracia@chromium.org>
3125
3126         Update the description of LayoutTestController::textSurroundingElement
3127         https://bugs.webkit.org/show_bug.cgi?id=90620
3128
3129         Reviewed by Adam Barth.
3130
3131         Rename and fix the description of the method as its behaviour changed
3132         during the review of https://bugs.webkit.org/show_bug.cgi?id=82461
3133
3134         * DumpRenderTree/chromium/LayoutTestController.cpp:
3135         (LayoutTestController::LayoutTestController):
3136         (LayoutTestController::textSurroundingNode):
3137         * DumpRenderTree/chromium/LayoutTestController.h:
3138         (LayoutTestController):
3139
3140 2012-07-08  Kent Tamura  <tkent@chromium.org>
3141
3142         Improve the template generated by prepare-ChangeLog
3143         https://bugs.webkit.org/show_bug.cgi?id=89560
3144
3145         Reviewed by Ryosuke Niwa.
3146
3147         Produce the following template:
3148
3149         > 2012-06-20  Kent Tamura  <tkent@chromium.org>
3150         >
3151         >         Need a short description (Oops!).
3152         >         Need the bug URL (Oops!).
3153         >
3154         >         Reviewed by Ryosuke Niwa.
3155         >
3156         >         Additional information of the change such as approach, rationale. Please add per-function descriptions below. (Oops!).
3157         >
3158         >         No new tests (Oops!).
3159
3160         * Scripts/prepare-ChangeLog:
3161         (generateNewChangeLogs):
3162
3163 2012-07-08  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
3164
3165         [wx] Unreviewed. Fix notificaitons directory path after notifications move.
3166
3167         * waf/build/settings.py:
3168
3169 2012-07-07  Nuno Lopes  <nlopes@apple.com>
3170
3171         Fix build with recent clang.
3172         https://bugs.webkit.org/show_bug.cgi?id=90712
3173
3174         Reviewed by Alexey Proskuryakov.
3175
3176         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
3177         (DOMWindowExtensionBasic):
3178         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp:
3179         (DOMWindowExtensionNoCache):
3180         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash_Bundle.cpp:
3181         (TestWebKitAPI::DocumentStartUserScriptAlertCrashTest::initialize):
3182
3183 2012-07-06  Zan Dobersek  <zandobersek@gmail.com>
3184
3185         [Gtk] Add a configuration option for disabling unstable features in releases
3186         https://bugs.webkit.org/show_bug.cgi?id=87995
3187
3188         Reviewed by Martin Robinson.
3189
3190         Pass the --enable-unstable-features flag when building the Gtk port through build-webkit.
3191
3192         * Scripts/webkitdirs.pm:
3193         (buildAutotoolsProject):
3194
3195 2012-07-06  Jessie Berlin  <jberlin@apple.com>
3196
3197         WKContext should ask for its initialization data when it needs it so the client doesn't have
3198         to keep it up to date.
3199         https://bugs.webkit.org/show_bug.cgi?id=90627
3200
3201         Reviewed by Anders Carlsson.
3202
3203         Add tests and update other WKContextInjectedBundleClients.
3204
3205         * MiniBrowser/mac/AppDelegate.m:
3206         (-[BrowserAppDelegate init]):
3207         Updated for the change to WKContextInjectedBundleClient.
3208         * WebKitTestRunner/TestController.cpp:
3209         (WTR::TestController::initialize):
3210         Ditto.
3211
3212         * TestWebKitAPI/PlatformUtilities.cpp:
3213         (TestWebKitAPI::Util::createInitializationDictionaryForInjectedBundleTest):
3214         Moved the logic to create the initialization dictionary here ...
3215         (TestWebKitAPI::Util::createContextForInjectedBundleTest):
3216         ... from here so that it can be used without automatically using
3217         WKContextSetInitializationUserDataForInjectedBundle.
3218         * TestWebKitAPI/PlatformUtilities.h:
3219
3220         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3221         Add the new test files.
3222         * TestWebKitAPI/GNUmakefile.am:
3223         Ditto.
3224
3225         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback.cpp: Added.
3226         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
3227         Check that the message received from the injected bundle matches the user data it was
3228         initialized with.
3229         (TestWebKitAPI::getInjectedBundleInitializationUserData):
3230         Return the user data that the injected bundle should be initialized with.
3231         (TestWebKitAPI::TEST):
3232         Set up WKContextInjectedBundleClient and load a page.
3233         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback_Bundle.cpp: Added.
3234         (TestWebKitAPI::GetInjectedBundleInitializationUserDataCallbackTest::initialize):
3235         Send the initialization user data back up to the UI Process.
3236
3237         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins.cpp: Added.
3238         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
3239         Check that the message received from the injected bundle matches the user data it was
3240         initialized with in the callback.
3241         (TestWebKitAPI::getInjectedBundleInitializationUserData):
3242         Return the user data that the injected bundle should be initialized with.
3243         (TestWebKitAPI::TEST):
3244         Set up the context and use WKContextSetInitializationUserDataForInjectedBundle to set the
3245         initialization user data (which should be overridden by the user data returned in
3246         getInjectedBundleInitializationUserData).
3247         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins_Bundle.cpp: Added.
3248         (TestWebKitAPI::InjectedBundleInitializationUserDataCallbackWinsTest::initialize):
3249         Send the initialization user data back up to the UI Process.
3250
3251 2012-07-06  Dongwoo Im  <dw.im@samsung.com>
3252
3253         [EFL][GTK] jhbuild : Disable pixman demos build depending on GTK+
3254         https://bugs.webkit.org/show_bug.cgi?id=90593
3255
3256         Reviewed by Philippe Normand.
3257
3258         * efl/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
3259         * gtk/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
3260
3261 2012-07-06  Luciano Wolf <luciano.wolf@openbossa.org>
3262
3263         [Qt] DumpRenderTree does not use 'monospace' font when directed
3264         https://bugs.webkit.org/show_bug.cgi?id=85203
3265
3266         The test fonts used for Qt tests were moved to the Liberation font family.
3267
3268         Due to this change we are skipping tons of tests. They will be gradually
3269         unskipped in batches, ASAP.
3270
3271         Also, the now unneeded special font configuration for Qt5 WK1 was removed.
3272
3273         Reviewed by Tor Arne Vestbø.
3274
3275         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
3276         (WebKit::initializeTestFonts):
3277
3278 2012-07-06  Csaba Osztrogonác  <ossy@webkit.org>
3279
3280         [Qt] Buildfix for newer Qt5
3281         https://bugs.webkit.org/show_bug.cgi?id=90519
3282
3283         Reviewed by Tor Arne Vestbø.
3284
3285         * MiniBrowser/qt/MiniBrowser.pro: We also need to depend on gui-private in the MiniBrowser to get access to these headers.
3286         * MiniBrowser/qt/MiniBrowserApplication.h: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
3287         * WebKitTestRunner/qt/PlatformWebViewQt.cpp: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
3288
3289 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3290
3291         [Qt] Make use of Qt5 qmake's changed makefile recursion behavior
3292
3293         When not using the -r option, qmake now interleaves qmake and make calls,
3294         so we don't need custom logic for this. There's also an option to supress
3295         the effect of the -r option, which we use to make WebKit.pro the only
3296         project file parsed in a recursive qmake-run.
3297
3298         https://bugs.webkit.org/show_bug.cgi?id=90461
3299
3300         Reviewed by Tor Arne Vestbø.
3301
3302         * qmake/mkspecs/features/default_post.prf:
3303         * qmake/mkspecs/features/functions.prf:
3304
3305 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3306
3307         [Qt] Remove custom qmake logic for module creation
3308
3309         Qmake now has the necessary hooks to cleanly override the build locations.
3310
3311         https://bugs.webkit.org/show_bug.cgi?id=90461
3312
3313         Reviewed by Tor Arne Vestbø.
3314
3315         * qmake/mkspecs/features/default_pre.prf:
3316
3317 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3318
3319         [Qt] Add top-level .qmake.conf
3320
3321         With Qt5, this makes setting $QMAKEPATH externally unnecessary.
3322
3323         The magic in the perl scripts to set QMAKEPATH is still there, as it doesn't
3324         hurt, and is still required for Qt4.
3325
3326         https://bugs.webkit.org/show_bug.cgi?id=90461
3327
3328         Reviewed by Tor Arne Vestbø.
3329
3330         * qmake/mkspecs/features/default_pre.prf:
3331
3332 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3333
3334         [Qt] Unify qtFeatureDefaults code paths
3335
3336         There's no reason to run qmake on features.prf directly anymore, as we
3337         can selectivly run configure tests in the project file now.
3338
3339         https://bugs.webkit.org/show_bug.cgi?id=90461
3340
3341         Reviewed by Tor Arne Vestbø.
3342
3343         * Scripts/webkitdirs.pm:
3344         (qtFeatureDefaults):
3345         * qmake/configure.pri:
3346         * qmake/mkspecs/features/features.prf:
3347
3348 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3349
3350         [Qt] Don't let qt_webkit.pri proclaim its own location
3351
3352         This won't work any more with recent Qt5 versions, as the forwarding
3353         pri is created by qt_module_config, which needs MODULE_PRI to be set
3354         up already.
3355
3356         We also need to load build_config, not qt_module.
3357
3358         https://bugs.webkit.org/show_bug.cgi?id=90461
3359
3360         Reviewed by Tor Arne Vestbø.
3361
3362         * qmake/qt_webkit.pri:
3363
3364 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3365
3366         [Qt] Remove redundant CONFIG+=module
3367
3368         qt_module_config takes care of that.
3369
3370         In api.pri we are actually testing the flag ourselves, so now we need to
3371         test a related flag qt_module_config sets instead.
3372
3373         Reviewed by Tor Arne Vestbø.
3374
3375         * qmake/mkspecs/features/default_post.prf:
3376
3377 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
3378
3379         [Qt] Switch to new-style Qt 5 configure tests
3380
3381         Use explicit project file action instead of syncqt magic.
3382
3383         https://bugs.webkit.org/show_bug.cgi?id=90461
3384
3385         Reviewed by Tor Arne Vestbø.
3386
3387         * Scripts/webkitdirs.pm:
3388         (buildQMakeProjects):
3389         * qmake/configure.pri: Copied from Tools/qmake/configure.pro.
3390         * qmake/configure.pro:
3391         * qmake/mkspecs/features/default_pre.prf:
3392         * qmake/mkspecs/features/features.prf:
3393         * qmake/sync.profile: Removed.
3394
3395 2012-07-05  Jason Liu  <jason.liu@torchmobile.com.cn>
3396
3397         Add Jason Liu to committers.py.
3398         https://bugs.webkit.org/show_bug.cgi?id=90654
3399
3400         Unreviewed.
3401
3402         * Scripts/webkitpy/common/config/committers.py:
3403
3404 2012-07-05  Benjamin Poulain  <bpoulain@apple.com>
3405
3406         Double release of resources if the load is canceled in a callback of ResourceLoader::didFinishLoading
3407         https://bugs.webkit.org/show_bug.cgi?id=90431
3408
3409         Reviewed by Anders Carlsson.
3410
3411         Add a Mac API test.
3412
3413         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3414         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.html: Added.
3415         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.mm: Added.
3416         (-[CancelLoadFromResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
3417         (-[CancelLoadFromResourceLoadDelegateFrameLoadDelegate webView:didFinishLoadForFrame:]):
3418         (TestWebKitAPI):
3419         (TestWebKitAPI::TEST):
3420
3421 2012-07-05  Dave Tharp  <dtharp@codeaurora.org>
3422
3423         Adding myself as committer to committers.py
3424
3425         Unreviewed.
3426
3427         * Scripts/webkitpy/common/config/committers.py:
3428
3429 2012-07-05  Alexey Proskuryakov  <ap@apple.com>
3430
3431         [Mac][WK2] Enable HTTPS tests
3432         https://bugs.webkit.org/show_bug.cgi?id=90600
3433
3434         Reviewed by Brady Eidson.
3435
3436         * DumpRenderTree/mac/DumpRenderTree.mm: (dumpRenderTree): Removed a misleading comment.
3437         It explained why we were doing this here instead of a frame load delegate, but that's
3438         just history, not an explanation of why we need this code.
3439
3440         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: (WTR::InjectedBundle::platformInitialize):
3441         Ignore certificate errors for localhost and 127.0.0.1. Since this is only needed
3442         for tests, it seems appropriate to have platform specific solutions in every WTR,
3443         and not add a WebKit2 API.
3444
3445 2012-07-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3446
3447         [Qt] Remove rpath-hacks in qt_webkit.pri
3448
3449         Having it in qt_webkit.pri meant that all client apps would inherit the rpath,
3450         even if they were not using QtWebKit.
3451
3452         This is a regression for people who build client apps against a non-installed
3453         QtWebKit, as the client app will be missing a rpath to the standalone WebKit
3454         build dir. This can be solved by LD_LIBRARY_PATH or custom QMAKE_RPATHDIR
3455         logic, until the real issue is fixed in Qt.
3456
3457         Reviewed by Csaba Osztrogonác.
3458
3459         * qmake/mkspecs/features/rpath.prf:
3460         * qmake/qt_webkit.pri:
3461
3462 2012-07-05  Dongwoo Im  <dw.im@samsung.com>
3463
3464         [EFL] Enable the CUSTOM_SCHEME_HANDLER feature as default.
3465         https://bugs.webkit.org/show_bug.cgi?id=88608
3466
3467         Reviewed by Hajime Morita.
3468
3469         * Scripts/webkitperl/FeatureList.pm: Enable the CUSTOM_SCHEME_HANDLER feature as default.
3470
3471 2012-07-05  Ryuan Choi  <ryuan.choi@samsung.com>
3472
3473         [Wk2][EFL] EFL needs a WebKitTestRunner
3474         https://bugs.webkit.org/show_bug.cgi?id=87659
3475
3476         Reviewed by Kenneth Rohde Christiansen.
3477
3478         Add an implementation of WebKitTestRunner for Efl port.
3479
3480         * CMakeLists.txt: Includes WebKitTestRunner build script.
3481         * Scripts/build-webkittestrunner: Added knowledge of WebKitTestRunner/Efl.
3482         * Scripts/webkitpy/layout_tests/port/efl.py: Ditto.
3483         (EflPort.setup_environ_for_server):
3484         * WebKitTestRunner/CMakeLists.txt: Added.
3485         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3486         * WebKitTestRunner/InjectedBundle/efl/ActivateFontsEfl.cpp: Added.
3487         * WebKitTestRunner/InjectedBundle/efl/InjectedBundleEfl.cpp: Added.
3488         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp: Added.
3489         * WebKitTestRunner/PlatformEfl.cmake: Added.
3490         * WebKitTestRunner/PlatformWebView.h:
3491         * WebKitTestRunner/TestController.cpp:
3492         * WebKitTestRunner/config.h: Added knowledge of WebKitTestRunner/Efl.
3493         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp: Added.
3494         * WebKitTestRunner/efl/TestControllerEfl.cpp: Added.
3495         * WebKitTestRunner/efl/main.cpp: Added.
3496         (main):
3497
3498 2012-07-05  Christophe Dumez  <christophe.dumez@intel.com>
3499
3500         [WK2][EFL] Ewk_View needs API to load HTML data
3501         https://bugs.webkit.org/show_bug.cgi?id=90540
3502
3503         Reviewed by Kenneth Rohde Christiansen.
3504
3505         Update EFL MiniBrowser to catch the "load,error" signal
3506         on the view and display an error page.
3507
3508         * MiniBrowser/efl/main.c:
3509         (on_error):
3510         (browserCreate):
3511
3512 2012-07-05  Sergio Villar Senin  <svillar@igalia.com>
3513
3514         Hidden dirs are not copied when creating the built product archive
3515         https://bugs.webkit.org/show_bug.cgi?id=90559
3516
3517         Reviewed by Carlos Garcia Campos.
3518
3519         Hidden dirs are not copied inside the built product archive
3520         because the recursive copy command was ignoring them. We need those
3521         hidden directories for the GTK WebKit2 testing bot to work fine.
3522
3523         * BuildSlaveSupport/built-product-archive:
3524         (archiveBuiltProduct): replaced "*" by "." to include hidden
3525         dirs/files.
3526
3527 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
3528
3529         Unreviewed Chromium gardening, Disable always failed python tests on Windows.
3530         https://bugs.webkit.org/show_bug.cgi?id=90587
3531
3532         PortTest.qt_linux and mac_lion are always failed on Chromium Windows at
3533         creating child process.
3534
3535         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3536         (PortTest.disabled_test_qt_linux):
3537         (PortTest.disabled_test_mac_lion):
3538
3539 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
3540
3541         [Tools] webkit_unittest.py got assertion
3542         https://bugs.webkit.org/show_bug.cgi?id=90579
3543
3544         Reviewed by Hajime Morita.
3545
3546         This patch updates expectation of assertion in test_skipped_directories_for_symbols().
3547         r121874 added new (exclude) directory (inspector/styles/variables), however it didn't
3548         update one of two assertions.
3549
3550         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3551
3552 2012-07-04  Balazs Ankes  <bank@inf.u-szeged.hu>
3553
3554         webkit-patch should add reviewer if "Reviewed by NOBODY ..." is missing
3555         https://bugs.webkit.org/show_bug.cgi?id=67935
3556
3557         Reviewed by Ryosuke Niwa.
3558
3559         * Scripts/webkitpy/common/checkout/changelog.py:
3560         (ChangeLog.set_reviewer):
3561         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3562         (test_set_reviewer):
3563         (test_set_short_description_and_bug_url):
3564
3565 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
3566
3567         Unreviewed, rolling out r121862.
3568         http://trac.webkit.org/changeset/121862
3569         https://bugs.webkit.org/show_bug.cgi?id=90569
3570
3571         It broke built-product-archive for GTK, Qt and EFL (Requested
3572         by svillar on #webkit).
3573
3574         * BuildSlaveSupport/built-product-archive:
3575         (archiveBuiltProduct):
3576
3577 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
3578
3579         Hidden dirs are not copied when creating the built product archive
3580         https://bugs.webkit.org/show_bug.cgi?id=90559
3581
3582         Reviewed by Carlos Garcia Campos.
3583
3584         Hidden dirs are not copied inside the built product archive because
3585         the recursive copy command was ignoring them. We need those hidden
3586         directories for the GTK WebKit2 testing bot to work fine.
3587
3588         * BuildSlaveSupport/built-product-archive:
3589         (archiveBuiltProduct): removed the trailing "*" from the copy
3590         command.
3591
3592 2012-07-04  John Mellor  <johnme@chromium.org>
3593
3594         Text Autosizing: Add compile flag and runtime setting
3595         https://bugs.webkit.org/show_bug.cgi?id=87394
3596
3597         This patch renames Font Boosting to Text Autosizing.
3598
3599         Reviewed by Adam Barth.
3600
3601         * Scripts/webkitperl/FeatureList.pm:
3602         * qmake/mkspecs/features/features.pri:
3603
3604 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3605
3606         [Qt] Replace deprecated IN_PWD with PWD
3607
3608         Reviewed by Csaba Osztrogonác.
3609
3610         * qmake/mkspecs/features/default_pre.prf:
3611
3612 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3613
3614         [Qt] Don't try to build WTR twice
3615
3616         Reviewed by Csaba Osztrogonác.
3617
3618         * Tools.pro:
3619
3620 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
3621
3622         Unreviewed, rolling out r121821.
3623         http://trac.webkit.org/changeset/121821
3624         https://bugs.webkit.org/show_bug.cgi?id=90551
3625
3626         This patch did not receive a high-quality review and has a
3627         number of errors (Requested by abarth on #webkit).
3628
3629         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3630         (Build.results_url):
3631         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
3632         (ChromiumBuilder.accumulated_results_url):
3633         * Scripts/webkitpy/layout_tests/port/builders.py:
3634         * Scripts/webkitpy/tool/commands/rebaseline.py:
3635         (AbstractParallelRebaselineCommand.__init__):
3636         (Rebaseline):
3637         (Rebaseline._builder_to_pull_from):
3638         (Rebaseline._tests_to_update):
3639         (Rebaseline.execute):
3640         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3641         (test_rebaseline.mock_builder_to_pull_from):
3642         (test_rebaseline):
3643
3644 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3645
3646         [Qt] Get rid of un-needed QT += declarative for Qt 5
3647
3648         The declarative module has been renamed to quick1 in Qt 5, and the
3649         engine-only module for Qt 5 is named 'qml'. For many of the instances
3650         we could just remove 'declarative', since the project file was only
3651         used for Qt5/WebKit2 builds. In the other cases the module was wrapped
3652         in a haveQt(4) scope.
3653
3654         Reviewed by Csaba Osztrogonác.
3655
3656         * MiniBrowser/qt/MiniBrowser.pro:
3657         * WebKitTestRunner/InjectedBundle/Target.pri:
3658         * WebKitTestRunner/Target.pri:
3659
3660 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3661
3662         [Qt] Remove a few un-needed load(features) after r121777
3663
3664         The features are computed by configure.pro and cached in .qmake.cache.
3665
3666         Reviewed by Jocelyn Turcotte..
3667
3668         * MiniBrowser/qt/raw/Target.pri:
3669
3670 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3671
3672         [Qt] Use haveQtModule() in project files instead of manual checks
3673
3674         Reviewed by Jocelyn Turcotte.
3675
3676         * qmake/mkspecs/features/default_pre.prf:
3677         * qmake/mkspecs/features/features.prf:
3678
3679 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3680
3681         [Qt] Fix haveQtModule() check
3682
3683         Without {} brackets we were pulling out the value of the module.name variable.
3684
3685         Reviewed by Jocelyn Turcotte.
3686
3687         * qmake/mkspecs/features/functions.prf:
3688
3689 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
3690
3691         [WK2] [GTK] WK2 testing bot unable to find the InjectedBundle library
3692         https://bugs.webkit.org/show_bug.cgi?id=90541
3693
3694         Reviewed by Carlos Garcia Campos.
3695
3696         We must add the contents of WebKitBuild/*/Libraries/ directory to
3697         the file to be uploaded to b.w.o to successfully run WebKit2 tests
3698         in the WebKit2 testing bot.
3699
3700         * BuildSlaveSupport/built-product-archive:
3701         (archiveBuiltProduct): appended Libraries to the list of
3702         neededDirectories.
3703
3704 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
3705
3706         [GTK] WebKit2 testing bot unable to find the build directory
3707         https://bugs.webkit.org/show_bug.cgi?id=90523
3708
3709         Reviewed by Carlos Garcia Campos.
3710
3711         The presence of GNUmakefile was used to determine whether a given
3712         directory was a valid build directory or not. That's not correct for
3713         testing bots as we export only the minimum required stuff to run the
3714         tests.
3715
3716         * gtk/common.py:
3717         (get_build_path.is_valid_build_directory): added an extra check.:
3718
3719 2012-07-04  Joe Thomas  <joethomas@motorola.com>
3720
3721         Adding myself to committers list.
3722
3723         Unreviewed.
3724
3725         * Scripts/webkitpy/common/config/committers.py:
3726
3727 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
3728
3729         [WK2][EFL] Ewk_View should report the load progress
3730         https://bugs.webkit.org/show_bug.cgi?id=90457
3731
3732         Reviewed by Kenneth Rohde Christiansen.
3733
3734         Update EFL MiniBrowser so that it listens for the
3735         "load,progress" on the Ewk_View and updates its
3736         window title accordingly.
3737
3738         * MiniBrowser/efl/main.c:
3739         (title_set):
3740         (on_title_changed):
3741         (on_progress):
3742         (browserCreate):
3743
3744 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
3745
3746         disable failing mock-chromium-mac python unit tests
3747
3748         Unreviewed, build fix.
3749
3750         Disable these tests until I can figure out why they're failing
3751         on the bots.
3752
3753         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3754         (PortTest.disabled_test_chromium_mac_lion):
3755         (PortTest.disabled_test_chromium_mac_lion_in_test_shell_mode):
3756
3757 2012-07-03  Ian Vollick  <vollick@chromium.org>
3758
3759         Add vollick to the list of committers
3760         https://bugs.webkit.org/show_bug.cgi?id=90512
3761
3762         Reviewed by Adrienne Walker.
3763
3764         * Scripts/webkitpy/common/config/committers.py:
3765
3766 2012-07-03  Ojan Vafai  <ojan@chromium.org>
3767
3768         Improve webkit-patch rebaseline to work for more cases
3769         https://bugs.webkit.org/show_bug.cgi?id=90504
3770
3771         Reviewed by Dirk Pranke.
3772
3773         -Makes it work for the build.chromium.org bots.
3774         -Lets you rebaseline all builders instead of just one.
3775         -Lets you pass in the builder or test to rebaseline.
3776
3777         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3778         (Build.results_url):
3779         Make this work for build.chromium.org builders as well.
3780
3781         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
3782         (ChromiumBuilder):
3783         (ChromiumBuilder.latest_cached_build):
3784         ChromiumBuilder doesn't support large parts of the Builder interface.
3785         This provides the bare minimum for this patch to work. Eventually,
3786         we should create a single interface that can be supported for both
3787         sets of buildbots.
3788
3789         * Scripts/webkitpy/layout_tests/port/builders.py:
3790         (builder_path_from_name):
3791         (all_builder_names):
3792         memoizing here is incorrect because the test override _exact_matches,
3793         so these can return different values. In either case, I'm pretty sure these
3794         are not remotely hot codepaths.
3795
3796         * Scripts/webkitpy/tool/commands/rebaseline.py:
3797         (AbstractParallelRebaselineCommand.__init__):
3798         (Rebaseline):
3799         (Rebaseline.__init__):
3800         (Rebaseline._builders_to_pull_from):
3801         (Rebaseline._builder_with_name):
3802         (Rebaseline._tests_to_update):
3803         (Rebaseline.execute):
3804         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3805         (test_rebaseline.mock_builders_to_pull_from):
3806         (test_rebaseline):
3807         (test_rebaseline_command_line_flags):
3808         (test_rebaseline_multiple_builders):
3809         (test_rebaseline_multiple_builders.mock_builders_to_pull_from):
3810         (test_rebaseline_multiple_builders.mock_tests_to_update):
3811
3812 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
3813
3814         [EFL] Enable CSS variables support at compile time
3815         https://bugs.webkit.org/show_bug.cgi?id=90448
3816
3817         Reviewed by Kenneth Rohde Christiansen.
3818
3819         Turn on CSS_VARIABLES flag by default on EFL port.
3820
3821         * Scripts/webkitperl/FeatureList.pm:
3822
3823 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
3824
3825         nrwt: clean up exception handling and make sure we log some more failures
3826         https://bugs.webkit.org/show_bug.cgi?id=90503
3827
3828         Reviewed by Ojan Vafai.
3829
3830         There were several places where exceptions weren't getting
3831         logged, most notably if you passed a bad value to --platform.
3832         This change tests that and cleans things up a bit; more cleanup
3833         will be possible when we rework the manager_worker_broker code.
3834
3835         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3836         (_BrokerConnection.raise_exception):
3837         (_InlineWorkerConnection.raise_exception):
3838         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3839         (Worker.run):
3840         (Worker.kill_driver):
3841         * Scripts/webkitpy/layout_tests/port/factory.py:
3842         (PortFactory.get):
3843         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3844         (run):
3845         (main):
3846         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3847         (MainTest.test_unsupported_platfrom):
3848
3849 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
3850
3851         nrwt: fix mock port
3852         https://bugs.webkit.org/show_bug.cgi?id=90500
3853
3854         Reviewed by Ojan Vafai.
3855
3856         The MockDRT code was never updated when we switched the chromium
3857         ports to using "drt mode" by default. This change updates that
3858         code, fixes a typo in the chromium port that went undetected
3859         (default_test_timeout_ms -> default_timeout_ms), and adds tests
3860         that actually exercise some of the mock ports. These tests are
3861         useful in that they will exercise the port-specific code in an
3862         end-to-end-manner, but they are a bit slow for some reason (>1s
3863         each) that I need to look into.
3864
3865         * Scripts/webkitpy/layout_tests/port/chromium.py:
3866         (ChromiumDriver.stop):
3867         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3868         (ChromiumAndroidPort.default_timeout_ms):
3869         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
3870         (TestChromiumAndroidPort.test_default_timeout_ms):
3871         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3872         (main):
3873         (parse_options):
3874         (MockTestShell):
3875         (MockTestShell.output_for_test):
3876         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3877         (MockDRTTest.assertTest):
3878         (MockTestShellTest):
3879         (MockTestShellTest.make_drt):
3880         (MockTestShellTest.test_test_shell_parse_options):
3881         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3882         (PortTest):
3883         (PortTest.assert_mock_port_works):
3884         (PortTest.test_chromium_mac_lion):
3885         (PortTest.test_chromium_mac_lion_in_test_shell_mode):
3886         (PortTest.test_qt_linux):
3887         (PortTest.test_mac_lion):
3888
3889 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
3890
3891         nrwt: make the worker class stand alone with a cleaner interface
3892         https://bugs.webkit.org/show_bug.cgi?id=90409
3893
3894         Reviewed by Ojan Vafai.
3895
3896         Currently the Worker class derives from AbstractWorker, which is
3897         kind of crufty and awkward; it would be better if we did not
3898         rely on shared state.
3899
3900         This change changes that so that Worker derives from object, and
3901         exposes the following interface:
3902           __init__() - called in the manager process
3903           safe_init() - called in the worker process to initialize
3904             unpicklable state
3905           handle() - a single routine to handle all messages
3906           cleanup() - called so the worker can clean up
3907
3908         Also, all of the "administrative" messages that are handled by
3909         the worker (notification of start/stop/etc.) move into
3910         manager_worker_broker - this reduces worker.py to just handling
3911         the mechanics of actually running each test.
3912
3913         For the moment, we do this by creating Yet Another wrapper/proxy
3914         class in manager_worker_broker, but this will get simpler
3915         shortly when the rest of m_w_b is cleaned up.
3916
3917         With this change worker is now in its new form but there will be
3918         a follow-on change that cleans up some names and other minor
3919         things.
3920
3921         This change is again mostly just moving things around and should
3922         be covered by the (updated) existing tests.
3923
3924         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3925         (get):
3926         (AbstractWorker.__init__):
3927         (AbstractWorker.run):
3928         (AbstractWorker):
3929         (AbstractWorker.handle_stop):
3930         (AbstractWorker.handle_test_list):
3931         (AbstractWorker.yield_to_broker):
3932         (AbstractWorker.post_message):
3933         (_WorkerConnection.__init__):
3934         (_Process.run):
3935         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
3936         (_TestWorker):
3937         (_TestWorker.__init__):
3938         (_TestWorker.name):
3939         (_TestWorker.cleanup):
3940         (_TestWorker.handle):
3941         (_TestWorker.safe_init):
3942         (_TestWorker.stop):
3943         (_TestsMixin.handle_finished_test):
3944         (_TestsMixin.setUp):
3945         (_TestsMixin.test_cancel):
3946         (_TestsMixin.test_done):
3947         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3948         (Worker):
3949         (Worker.__init__):
3950         (Worker.safe_init):
3951         (Worker.handle):
3952
3953 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
3954
3955         nrwt: moving child process logging code into manager_worker_broker
3956         https://bugs.webkit.org/show_bug.cgi?id=90408
3957
3958         Reviewed by Ojan Vafai.
3959
3960         Users of manager_worker_broker should not have to be aware of
3961         whether they're in the same process or different processes and
3962         configure logging themselves; mwb should hide this complexity.
3963         We can't quite do this completely/correctly yet, since the
3964         manager expects to get a list of messages to log, but this
3965         change fixes the worker side of it, at least.
3966
3967         This is just moving code around, there is no new functionality
3968         and this should be covered by existing tests.
3969
3970         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3971         (AbstractWorker.__init__):
3972         (_WorkerConnection.__init__):
3973         (_WorkerConnection.post_message):
3974         (_WorkerConnection):
3975         (_WorkerConnection.set_up_logging):
3976         (_WorkerConnection.clean_up_logging):
3977         (_InlineWorkerConnection.run):
3978         (_MultiProcessWorkerConnection.run):
3979         (_WorkerLogHandler):
3980         (_WorkerLogHandler.__init__):
3981         (_WorkerLogHandler.emit):
3982         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
3983         (_TestWorker.run):
3984         (_TestsMixin.handle_done):
3985         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3986         (Worker.__init__):
3987         (Worker.run):
3988         (Worker._run_test):
3989         (Worker.cleanup):
3990         (Worker.run_single_test):
3991
3992 2012-07-03  Tony Chang  <tony@chromium.org>
3993
3994         [chromium] Don't archive build files generated by VS2010
3995         https://bugs.webkit.org/show_bug.cgi?id=90493
3996
3997         Reviewed by Ojan Vafai.
3998
3999         In particular, the pch (pre compiled header) files are gigantic, making
4000         release.zip almost a 1gb in size.
4001
4002         * BuildSlaveSupport/built-product-archive:
4003         (archiveBuiltProduct):
4004
4005 2012-07-03  Tony Chang  <tony@chromium.org>
4006
4007         [chromium] Unreviewed, update .gitignore to handle VS2010 files.
4008
4009         * .gitignore:
4010         * DumpRenderTree/DumpRenderTree.gyp/.gitignore:
4011         * TestWebKitAPI/TestWebKitAPI.gyp/.gitignore:
4012
4013 2012-07-03  Ojan Vafai  <ojan@chromium.org>
4014
4015         Have webkit-patch rebaseline use rebaseline-test-internal
4016         https://bugs.webkit.org/show_bug.cgi?id=90491
4017
4018         Reviewed by Dirk Pranke.
4019
4020         This lets it handle new files, reduces duplicate code and lays the
4021         groundwork for making rebaseline have a richer interface.
4022
4023         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
4024         (MockBuild):
4025         (MockBuild.__init__):
4026         (MockBuilder.build):
4027         * Scripts/webkitpy/tool/commands/rebaseline.py:
4028         (AbstractParallelRebaselineCommand._files_to_add):
4029         (Rebaseline):
4030         (Rebaseline._builder_to_pull_from):
4031         (Rebaseline._tests_to_update):
4032         (Rebaseline.execute):
4033         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4034         (test_overrides_are_included_correctly):
4035         (test_rebaseline):
4036         (test_rebaseline.mock_builder_to_pull_from):
4037         (test_rebaseline.mock_tests_to_update):
4038
4039 2012-07-03  Ojan Vafai  <ojan@chromium.org>
4040
4041         Rename rebaseline-test to rebaseline-test-internal
4042         https://bugs.webkit.org/show_bug.cgi?id=90485
4043
4044         Reviewed by Adam Barth.
4045
4046         It's now only used by other rebaseline commands. It's still useful to leave it
4047         as it's own command to aid in debugging when something goes wrong. In a followup
4048         patch, I'll make webkit-patch rebaseline cover any use-cases that rebaseline-test
4049         might have covered.
4050
4051         We no longer need the --print-scm-changes option since the only caller always passes
4052         that option in.
4053
4054         Also, make all the arguments command-line flags instead. Simplifies the code
4055         a bit in my opinion.
4056
4057         * Scripts/webkitpy/tool/commands/rebaseline.py:
4058         (RebaselineTest):
4059         (RebaselineTest.__init__):
4060         (RebaselineTest.execute):
4061         (AbstractParallelRebaselineCommand._rebaseline_commands):
4062         (RebaselineExpectations.execute):
4063         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4064         (test_rebaseline_all):
4065
4066 2012-07-03  Balazs Kelemen  <kbalazs@webkit.org>
4067
4068         [Qt][WTR] Get rid of using DumpRenderTreeSupportQt
4069         https://bugs.webkit.org/show_bug.cgi?id=90262
4070
4071         Unreviewed build fix.
4072
4073         Reverting r121550 manually because it caused build
4074         break on ARM and Mac. It turned out that the Qt build
4075         system is too broken for this change now, first we have
4076         to address two issues:
4077             * symbol hiding was forbidden accidentally on x86 Linux
4078             * Qt lacks a separate library for test support
4079
4080         * Tools.pro:
4081         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
4082         (WTR::InjectedBundlePage::resetAfterTest):
4083         (WTR::InjectedBundlePage::didClearWindowForFrame):
4084         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
4085
4086 2012-07-03  Tony Chang  <tony@chromium.org>
4087
4088         [chromium] Handle trailing backslash on %VSINSTALLDIR%
4089         https://bugs.webkit.org/show_bug.cgi?id=90410
4090
4091         Reviewed by Ojan Vafai.
4092
4093         * Scripts/webkitdirs.pm:
4094         (buildChromiumVisualStudioProject): Handle VSINSTALLDIR ending in a backslash, the backslash escapes the closing double quote.
4095
4096 2012-07-02  Ojan Vafai  <ojan@chromium.org>
4097
4098         Make the skia_test_expectations.txt file optional.
4099         https://bugs.webkit.org/show_bug.cgi?id=90400
4100
4101         Reviewed by Dirk Pranke.
4102
4103         It used to be optional. This regressed at some point. It's important that it be
4104         optional so that webkit-patch commands work in a pure-webkit checkout for chromium bots.
4105         Specifically, this was breaking webkit-patch rebaseline-test when it would go to update
4106         TestExpectations.
4107
4108         * Scripts/webkitpy/layout_tests/port/chromium.py:
4109         (ChromiumPort.expectations_files):
4110         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
4111         (ChromiumDriverTest.test_expectations_dict):
4112
4113 2012-07-03  Raphael Kubo da Costa  <rakuco@webkit.org>
4114
4115         [jhbuild][EFL] Bump libffi dependency.
4116         https://bugs.webkit.org/show_bug.cgi?id=85373
4117
4118         Reviewed by Gustavo Noronha Silva.
4119
4120         Update libffi to 3.0.11, which fixes the build on some platforms
4121         where 3.0.10 had problems.
4122
4123         * efl/jhbuild.modules:
4124
4125 2012-07-03  George Staikos  <staikos@webkit.org>
4126
4127         [BlackBerry] Enable microdata support for BlackBerry.
4128         https://bugs.webkit.org/show_bug.cgi?id=90429
4129
4130         Reviewed by Rob Buis.
4131
4132         * Scripts/webkitperl/FeatureList.pm:
4133
4134 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4135
4136         [Qt] Make use of .qmake.cache for caching features
4137
4138         Instead of loading() features from the files that need them (and re-running
4139         a bunch of checks), we now run feature detection as part of configure.pro,
4140         and have build-webkit write the computed feature-defines and CONFIG to
4141         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
4142         when building WebKit.pro.
4143
4144         At some point we'll be able to selectivly prevent running of config tests
4145         in configure.pro, which means we don't need a separate code-path for
4146         the build-webkit --help case.
4147
4148         We should also move the code in build-webkit that now uses .webkit.config
4149         to detect clean builds, to use .qmake.cache, since we now store the same
4150         thing there.
4151
4152         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
4153
4154         Reviewed by Tor Arne Vestbø.
4155
4156         * DumpRenderTree/qt/DumpRenderTree.pro:
4157         * QtTestBrowser/QtTestBrowser.pro:
4158         * Scripts/webkitdirs.pm:
4159         (qtFeatureDefaults):
4160         (buildQMakeProjects):
4161         * Tools.pro:
4162         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
4163         * WebKitTestRunner/InjectedBundle/Target.pri:
4164         * WebKitTestRunner/Target.pri:
4165         * qmake/.qmake.conf: Added.
4166         * qmake/configure.pro:
4167         * qmake/mkspecs/features/default_post.prf:
4168         * qmake/mkspecs/features/default_pre.prf:
4169         * qmake/mkspecs/features/features.prf:
4170
4171 2012-07-03  Nico Weber  <thakis@chromium.org>
4172
4173         -Wunused-private-field cleanup, linux edition
4174         https://bugs.webkit.org/show_bug.cgi?id=90463
4175
4176         Reviewed by Stephen White.
4177
4178         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
4179         * DumpRenderTree/chromium/GamepadController.h:
4180         (GamepadController):
4181         * DumpRenderTree/chromium/NotificationPresenter.h:
4182         (NotificationPresenter::NotificationPresenter):
4183         (NotificationPresenter):
4184
4185 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
4186
4187         Unreviewed, rolling out r121766.
4188         http://trac.webkit.org/changeset/121766
4189         https://bugs.webkit.org/show_bug.cgi?id=90465
4190
4191         It caused flakey build errors on the bots (Requested by Ossy
4192         on #webkit).
4193
4194         * DumpRenderTree/qt/DumpRenderTree.pro:
4195         * QtTestBrowser/QtTestBrowser.pro:
4196         * Scripts/webkitdirs.pm:
4197         (qtFeatureDefaults):
4198         (buildQMakeProjects):
4199         * Tools.pro:
4200         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
4201         * WebKitTestRunner/InjectedBundle/Target.pri:
4202         * WebKitTestRunner/Target.pri:
4203         * qmake/.qmake.conf: Removed.
4204         * qmake/configure.pro:
4205         * qmake/mkspecs/features/default_post.prf:
4206         * qmake/mkspecs/features/default_pre.prf:
4207         * qmake/mkspecs/features/features.prf:
4208
4209 2012-07-03  George Staikos  <staikos@webkit.org>
4210
4211         [BlackBerry] Enable Custom Scheme Handlers for BlackBerry.
4212         https://bugs.webkit.org/show_bug.cgi?id=90422
4213
4214         Reviewed by Rob Buis.
4215
4216         * Scripts/webkitperl/FeatureList.pm:
4217
4218 2012-07-03  George Staikos  <staikos@webkit.org>
4219
4220         [BlackBerry] Enable RegisterProtocolHandler for BlackBerry.
4221         https://bugs.webkit.org/show_bug.cgi?id=90422
4222
4223         Reviewed by Rob Buis.
4224
4225         * Scripts/webkitperl/FeatureList.pm:
4226
4227 2012-07-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
4228
4229         [EFL] Add run API tests as a step on the build bots
4230         https://bugs.webkit.org/show_bug.cgi?id=87252
4231
4232         Reviewed by Daniel Bates.
4233
4234         Run EFL API test suite on the build bots.
4235
4236         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4237         (RunEflAPITests):
4238         (TestFactory.__init__):
4239         (BuildAndTestFactory.__init__):
4240         * Scripts/run-efl-tests: Added.
4241
4242 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4243
4244         [Qt] Make use of .qmake.cache for caching features
4245
4246         Instead of loading() features from the files that need them (and re-running
4247         a bunch of checks), we now run feature detection as part of configure.pro,
4248         and have build-webkit write the computed feature-defines and CONFIG to
4249         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
4250         when building WebKit.pro.
4251
4252         At some point we'll be able to selectivly prevent running of config tests
4253         in configure.pro, which means we don't need a separate code-path for
4254         the build-webkit --help case.
4255
4256         We should also move the code in build-webkit that now uses .webkit.config
4257         to detect clean builds, to use .qmake.cache, since we now store the same
4258         thing there.
4259
4260         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
4261
4262         Reviewed by Tor Arne Vestbø.
4263
4264         * DumpRenderTree/qt/DumpRenderTree.pro:
4265         * QtTestBrowser/QtTestBrowser.pro:
4266         * Scripts/webkitdirs.pm:
4267         (qtFeatureDefaults):
4268         (buildQMakeProjects):
4269         * Tools.pro:
4270         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
4271         * WebKitTestRunner/InjectedBundle/Target.pri:
4272         * WebKitTestRunner/Target.pri:
4273         * qmake/.qmake.conf: Added.
4274         * qmake/configure.pro:
4275         * qmake/mkspecs/features/default_post.prf:
4276         * qmake/mkspecs/features/default_pre.prf:
4277         * qmake/mkspecs/features/features.prf:
4278
4279 2012-07-03  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>  Joel Dillon <joel.dillon@codethink.co.uk>
4280
4281         [Qt][Win] Fix broken QtWebKit5.lib linking
4282         https://bugs.webkit.org/show_bug.cgi?id=88321
4283
4284         Reviewed by Kenneth Rohde Christiansen.
4285
4286         On windows the import/export definition is part of the symbol's signature.
4287         Define STATICALLY_LINKED_WITH_$$library for each dependend module
4288         that is being linked statically to make sure that they can be linked together
4289         later on.
4290
4291         Also do not compile Assertions.cpp in DumpRenderTree anymore since all the
4292         used symbols are exported and it would cause a duplicate symbols error.
4293
4294         * DumpRenderTree/qt/DumpRenderTree.pro:
4295         * qmake/mkspecs/features/default_post.prf:
4296
4297 2012-07-03  Szilard Ledan  <szledan@inf.u-szeged.hu>
4298
4299         master.cfg cleanup, pass BuildStep instances instead of BuildStep subclasses
4300         https://bugs.webkit.org/show_bug.cgi?id=89001
4301
4302         Reviewed by Csaba Osztrogonác.
4303
4304         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4305         (UploadBuiltProduct.__init__):
4306         (UploadTestResults.__init__):
4307         (ExtractTestResults.__init__):
4308         (Factory.__init__):
4309         (BuildFactory.__init__):
4310         (TestFactory.__init__):
4311         (BuildAndTestFactory.__init__):
4312         (BuildAndPerfTestFactory.__init__):
4313         (BuildAndPerfTestWebKit2Factory.__init__):
4314         (DownloadAndPerfTestFactory.__init__):
4315
4316 2012-07-03  Ádám Kallai  <kadam@inf.u-szeged.hu>
4317
4318         Update unit test after r121724.
4319         https://bugs.webkit.org/show_bug.cgi?id=90444
4320
4321         Reviewed by Csaba Osztrogonác.
4322
4323         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
4324         (GardeningServerTest.test_rebaseline_new_port):
4325
4326 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
4327